College Student Attrition and Retention - Research
College Board Report No.Sl-1
College Entrance Examination Board, New York, 1981
Leonard Ramist is a. staff member of the Educational Testing Service, Princeton, New Jersey.
Researchers are encouraged to express freely their
professional judgment. Therefore, points of view or
opinions stated in College Board Reports do not
necessarily represent official College Board position
The College Board is a nonprofit membership organization that provides tests and other educational services
for students, schools, and colleges. The membership is composed of more than 2,500 colleges, schools, school
systems, and education associations. Representatives of the members serve on the Board of Trustees and advisory councils and committees that consider the programs of the College Board and participate in the determination of its policies and activities.
Copyright © 1981 by College Entrance Examination Board. All rights reserved.
Printed in the United States of America.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Overall Graduation and Dropout Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Students' Reasons for Dropping Out . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Straight Through College, Stopout, Dropout-What Difference Does It Make? . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Effects of College . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Leave of Absence ..................................................................... 5
Student Demographic Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Student Academic Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
StudentMotivationalFactors ............................................................... 10
Student Personal Characteristic Factors ....................................................... 11
General College Environmental Factors ....................................................... 12
College Programs and their Effects ........................................................... 13
Preenrollmentinformation ............................................................. 13
Admissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
College Costs and Ways to Meet Them ...................................................... 15
Orientation .......................................................................... 16
Faculty-Student Interaction ............................................................. 17
Academic Programs ................................................................... 18
Counseling and Advising ............................................................... 18
Career Development ................................................................... 20
Campus Activities ..................................................................... 21
Housing ............................................................................ 21
The Withdrawal Procedure .............................................................. 22
Two-Year Colleges .................................................................... 22
Adm1nistration of the Retention Program ................................................... 23
Summary .............................................................................. 24
In Conclusion ........................................................................... 30
References ............................................................................. 31
Students who drop out of college often suffer personal disappointments, financial setbacks, and a lowering of
career and life goals. Concern about the student has led to much research on college student attrition and
retention. This research has typically been in terms of statistical analyses of the differences between dropouts
and persisters. The reason for the research usually was to understand the phenomenon of dropping out,
which was assumed to be a major societal problem.
In the past few years, recognition of the imminent leveling off and decline of the number of students of
college-going age has lent a certain sense of institutional urgency not only to the understanding of which
students drop out and why, but also to influencing them to stay. In addition to its role in describing a societal
ill, the research on the causes of attrition has now taken on the dimension of an urgent administrative
necessity to keep students.
At the same time, a new movement for student consumer rights has developed. This movement has
challenged long-accepted practices and assumptions about higher education. One assumption that has been
challenged is that it is always in the best interest of the student to continue his or her education without inter·
ruption. Many students question the placement of any stigma on dropping or stopping out of college. They say
that information about options other than four consecutive years of college education should be made
available to them, and that administrative procedures for delayed entry after high school, for withdrawal, and
for reentry should be made as clear and efficient as possible. This position is consistent with the view of
education as a life-long learning process. Thus, while colleges are concerned about what they think are high
dropout rates and are gearing to combat this phenomenon, it is clear that any subtle societal stigma will no
longer be effective in retaining students.
But college concern about attrition and students' concern about their rights as consumers are really congruous. Although a great deal of sophisticated research has described the reasons for dropping out, with the
exception of the very few students who leave due to circumstances beyond their control, the student reasons
all stem from dissatisfaction with (or lack of sufficient perceived benefits from) the academic or social life of
the institution. For some, this dissatisfaction would be present regardless of the educational service that is
provided: they did not really want to go to college, the time is right for a break in their formal education, or
their talents could be better used elsewhere. For others, their original college choice was an error, and their
dissatisfaction could be eliminated by transfer. For still others, a better, more complete educational service offered by their college may have prevented or reduced the source of their dissatisfaction, may have changed
their perception of long-range college benefits, and may have resulted in retention. For the benefit of this latter group, and for the benefit of the perhaps larger group of dissatisfied students who would not consider
withdrawal as an option, a college can deal with its dropout problem by doing everything it can to upgrade the
educational service, in its broadest sense, that it provides its students. Such an attack on the root causes that
contribute to attrition would benefit all students and would be an excellent recruitment tactic for future
This book reviews research as it relates to college student attrition and retention. Overall dropout rates
and the reasons students give for dropping out are examined. Attempts are made to answer the series of questions: What difference does it make to go straight through college? Stopout? Dropout? Not go to college at all?
Examined next are the demographic, academic, motivational, and personal characteristics of the students who
are likely to drop out and how general college environmental factors are related to persistence. The concluding
sections describe college programs that would upgrade the level of educational service, thereby encouraging
students to stay.
OVERALL GRADUATION AND DROPOUT RATES
Given the multitude of dropout studies over the years, it may appear that it would be very easy to compute an
overall dropout rate. However, there are several problems related to definition.
(1) In some studies, transfers are considered dropouts and in others they are not. Transfers are dropouts
from the college of initial entry, not from higher education.
Dropout rates depend on the timing of the determination of dropout. Because a substantial proportion of students graduate in the fifth year after entry into a four-year college, a dropout rate calculated in terms of graduates after five years would be substantially lower than one calculated in
terms of graduates after four years. Indeed, the dropout rate calculated in terms of graduates after
ten years is still lower. To be sure of the permanency of a dropout-that there is not a later reentry
into higher education-each person would have to be tracked for the rest of his or her life.
Rates vary widely among different colleges. For example, a simple four-year nongraduation rate from
four-year colleges may vary from 10 percent to 85 percent (Summerskill, 1962; Noel, 1978). The
inclusion of students that originally entered two-year colleges has a particularly confounding effect
because, as expected, their rates of four-year degree completion are relatively low. Therefore, different samples of students from different colleges will yield different results.
In addition to the difficulties already present in comparing studies with different definitions, timing,
and samples, there may be real changes over time. That is, dropout rates and reasons for dropping
out may differ from one period of time to another.
Despite the difficulties, a careful review of past studies yields the tentative conclusions displayed in Table
1 about the overall graduation a.nd dropout rates of entrants to four-year colleges, based on a representative
cross section of four-year colleges.
Overall Graduation a.nd Dropout Rates
Graduation within four years after entry:
From original college of entry
From different college
Total - from a.ny college 4 years after entry
Graduation 5 years after entry (any college)
Graduation 6 or more years after entry (any college)
Total graduation (any college)
Dropouts who never receive degree
Several major studies have documented the 35-40 percent rate of graduation within four years after entry
from the college of entry shown in Table l. In a large study of 16,000 students from the 1931 and 1932 entering classes for the United States Department of Education, McNeeley (1937) found a. 38 percent graduation
rate. In a. similar study of 13,000 students from the 1950 entering class, Iffert (1958) found a. 39.5 percent
graduation rate. Summerskill (1962) found that the median rate for 35 studies done between 1920 and 1960
was 37 percent. Eckland (1964) found a. rate of 36.5 percent for the University of Illinois.
Three large longitudinal studies undertaken by the American Council on Education (ACE) substantiate the
45-60 percent total graduation rate within four years from a.ny college. They are identified in Table 2.
Graduation Rate Within Four Years from Any College
El-Khawa.s and Bisconti (1974, p. 60)
Bayer, Royer, and Webb (1973, p. 24)
Astin (1975, p. 11)
The latest evidence, in the National Longitudinal Study (NLS) of the class of 1972, conducted by the U.S.
Department of Health, Education, and Welfare, shows that 15 percent of the graduating high school class
finished a four-year college program within four years (National Center for Education Statistics, 1978, Volume
IV, p. 35). Given the 29 percent and 15 percent original entrance rates into four-year and two-yea.r colleges,
respectively (Fetters, 1977), and assuming a four-year degree completion rate of three to one between fouryear and two-year original college entry (found by Bayer, Royer, and Webb, 1973), an estimate of the comparable four-year graduation rate for the four-year college entering class of 1972 is only 45 percent. Although
still in the 45-60 percent range, it is possible that this rate has decreased substantially from the 1967 rate of
The fifth-year graduation rate of 10-15 percent is obtained from El-Khawas and Bisconti (1974, page 65).
They show a fifth-year rate of 14 percent for the class of 1961 and 13 percent for the class of 1966. Finally,
they show, for the class of 1961 only, a. siXth-year degree receipt rate of 4 percent; seventh-, eighth-, and
ninth-year rates of 2 percent, and a tenth-year rate of 1 percent. This total of 11 percent between the siXth and
tenth year for the class of 1961is the basis for 10-15 percent estimate of graduation six or more years of
study after entry.
The total graduation rate range of 65-90 percent implies a total dropout rate from higher education of
10-35 percent. Summersk.ill (1962) had ma.de a general statement tha.t this ra.te is about 40 percent. Although
his statement has often been quoted, it was made before the evidence provided by the recent ACE longitudinal
studies that there is indeed greater persistence (in its true sense) toward a degree several years after the
originally expected four-year graduation date tha.n he might ha.ve assumed.
The best single long-term graduation/dropout estimate is from the El-Khawas a.nd Bisconti (1974) 10-year
longitudinal data on the class of 1961: 80 percent of the four-year college entrants received a. degree within 10
years and 20 percent did not. The more recent NLS data on the class of 1972 ma.y show a higher dropout rate,
or after further follow-up evidence is in, it may merely show a higher stopout rate.
Of the 60-65 percent who do not graduate from their college of entry within four years, the interruption,
transfer, or termination is made in approximately equal percentages (of approximately 15 percent) in ea.ch of
the following four periods:
during the freshman year,
between the freshman year and the sophomore year,
during the sophomore year,
after the sophomore year.
Almost three-quarters of the students who complete their sophomore year without interruption complete all
four years without interruption.
STUDENTS' REASONS FOR DROPPING OUT
There is a large body of literature that is devoted to analyzing the reasons that students themselves give for
dropping out. In the process of summarizing the findings of over 100 research studies during a 25-yea.r
period, Pantages and Creedon (1978) categorize these reasons and show their relative frequency as follows:
Academic Matters. This most frequently cited category includes poor grades, boredom with courses, change
in career goals, and inability to take desired courses or programs. These reasons are given most often by men,
by those in technical institutes, by those with a poor high school. record, by those who drop out late in the
freshman year,a.nd by those who later transfer to another college or return. These reasons are given least by
those in teacher colleges or by those who drop out in the senior year. In contrasting Summerskill's (1962)
summary of several older studies with more recent studies, there appears to be a shift of emphasis from poor
grades to boredom with courses (Astin, 1975) and to the irrelevance of school work (Fetters, 1977).
Financial Difficulties. This second most frequently cited category is used primarily by minorities, women
with a poor high school record, early dropouts, and temporary dropouts. It is used least by those in technical
institutes or by women with a good high school record. Interestingly, counselors rate this category much
lower in importance than do students. Student reports of financial difficulties could be partially due to the fact
that financial reasons are more "socially acceptable."
Motivational Problems. This category Includes uncertainty about educational and occupational goals, lack of
interest in studies, and inability or unw1lllngness to study. It is cited particularly by early and permanent
dropouts, by academic dropouts, and by women with a poor high school record. It is rarely cited by women
with a good high school record. The fact that it is cited most often by academic dropouts could mean that a
lack of goal clarity contributes to low grades.
Personal Considerations. This category includes emotional problems, problems of adjustment to college life,
"getting one's head together," marriage, pregnancy, family responsibilities, and illness. Marriage is given as a
reason most often by women and minorities. This category is cited least by those in technical institutes.
Although students do not rate these reasons as high in importance, counselors indicate that socioemotional
problems are the most important reasons for attrition.
Dissatisfaction with the College. This category includes dissatisfaction with the size, the social or academic
environment, the regulations, etc. It Is cited most often by early and permanent dropouts.
Military Service. This category has become somewhat dated, but may take on more importance if the draft
is reinstated. It was cited by those who later reenrolled.
This reason is cited primarily by early dropouts and those in teacher colleges or two-year
Two other reasons reported by many students, which do not easily fit into the Pantages and Creedon
categories, are the expressed need for new, practical, nonacademic experiences and the lack of initial plans to
obtain a degree, particularly for two-year colleges.
STRAIGHT THROUGH COLLEGE, STOPOUT, DROPOUTWHAT DIFFERENCE DOES IT MAKE?
Effects of College
There is some important research documenting that college does have a significant effect on people's lives.
Withey (1971) found that college graduates are in general more optimistic and have better opportunities, more
job security, better working conditions, and higher job satisfaction. He found that they also tend to join more
organizations, take on more leadership roles, are better informed about national issues, and vote more often.
As shown in Table 3, the unemployment rate is directly related to the degree of educational attainment.
Extent of Education and Unemployment
Extent of Education
4 or more years of college
1-3 years of college
4 years of high school
1-3 years of high school
• United States Bureau of Labor Statistics data displayed
in United States Bureau of the Census, 1978, page 400.
Of course, it is not clear that college caused these outcomes. It could be that those who would more likely
experience these results seek out college.
As part of a landmark study on the determinants of economic success in America, Jencks ( 1977) studied
the financial effects of college. Even after controlling for family background and ability, Jencks and his colleagues found that college graduates enter higher status occupations than nongraduates and advance in status
faster. On a 0 to 96 scale of current occupational status, each year of college is worth 5 to 6 status points, and
degree completion is worth an additional 5 "bonus" status points-a total college effect of about 27 status
points (one standard deviation). A college dropout presumably would lose the 5 to 6 status points per uncompleted year plus the 5 degree completion status points. As a result of higher occupational status, college
graduates have about a 50 percent earnings advantage, which is reduced by about one-third if family
background and ability are controlled. These financial effects of college appear to be even more significant for
minorities and students from low-income families.
Based on a 10-year effort by the Cooperative Institutional Research Program of the American Council on
Education, which utilized longitudinal data from over 200,000 students and 300 institutions, Astin (1977)
reported findings on a major evaluation of the effects of the college experience. Longitudinal data collected
before and after college entry show that college students develop a more positive self-image, both in an interpersonal context and in terms of intellectual competence. They also develop more artistic interests, become
more liberal politically and socially, and become more competent in a great variety of areas. However, they
show marked declines in interest in traditional religious behavior, business, and altruism: they display an increase in hedonistic behavior.
Although some of these changes appear to be the result of maturation, Astin concluded that most appear
to be true effects of college. One piece of supporting evidence is that the college effects were not related to age.
Another is that the greater the student involvement on campus, the more pronounced were the college effects,
even after all other significant variables were controlled. For example, dropouts had a much smaller increase
in interpersonal self-esteem than did students who stayed in college, particularly those who were active in
fraternities or sororities.
However, another recent large study, part of the National Longitudinal Study of the class of 1972, offers
contrary evidence. Fetters (1977) reported that there was no significant difference in self-esteem in 1974 between students in college and those who had withdrawn from college in the previous two years. Perhaps it
takes longer than two years for these effects to become apparent.
In another large study of college effects, Trent and Medsker (1968) compared an employed group of high
school graduates with a group of students who spent the four years after high school in college. Even though
they controlled for academic ability and socioeconomic characteristics, they still found that, four years after
high school, the college group showed a three-times greater increase on a social maturity scale than did the
Some contrary evidence notwithstanding, we tentatively conclude that college does indeed make a difference in self-image, social maturity, interest, competencies, and employment.
Leave of Absence
Does a leave of absence alter the effects of college in any significant way? Pantages and Creedon (1978)
reviewed several longitudinal studies, which showed that about 70 percent of the students who leave a fouryear college eventually reenroll at some other college. NLS data (Fetters, 1977) show that about one-third of
the returning students return after one year. Of those returning, about 60 percent eventually earn a baccalaureate degree, a rate somewhat lower than the range of 65-90 percent for entering freshmen.
Cope and Hannah categorized the reenrollment and eventual graduation rates by type of institution as
shown in Table 4. Also, Haagen (1977) reported that for highly selective liberal arts colleges, over 90 percent
return, 80 percent within one year.
Rates of Return and Graduation, by Type of Institution •
Type of Institution •
Rate of Return
Rate of Graduation
Prestigious private universities
Better public universities
Typical state universities
Junior and community colleges
•cope and Hannah (1975, page 61). Private liberal arts colleges were purposely excluded because of their diversity.
After reviewing a large number of personal histories, Cope and Hannah (1975) identified some interesting
patterns of students who leave college:
-Students with high academic potential who fall tend to transfer to another college and do well.
-Upper middle class American students who question the value of college tend to reenroll after a year of
vagabonding, particularly in Europe.
-Students who suffer a severe disappointment in a relationship, particularly women, transfer to another
college, usually closer to home.
-Students who become disenchanted with the bureaucracy at a large university transfer to a small
-Students who wish to have the anonymity and excitement of a large university transfer to one.
-Students with a strong academic record and a lack of commitment to a degree find a job and do not give
college much further thought.
Haagen reported the activities in which students returning to selective liberal arts colleges spent a
substantial amount of time during their leave of absence, as shown in Table 6.
Activities During Leave of Absence•
Formal academic study
•Haagen (1977, page 65).
Of these students, 88 percent stated that their leave was positive, 4 percent that it was neutral, and only 8
percent that it was negative. Many had experiences relating to careers; others improved social skills by spending time outside an intellectual and analytical environment, in particular outside their student role; others
developed appreciation of different cultures, people, life-styles, and systems of values; and others dealt with
psychological problems that may have hampered their academic performance. The great majority felt more
self-confident and independent, less timid, and more effective in relating to faculty. After their return, they
were less dependent on faculty advice, but more insistent on faculty intellectual leadership.
Fetters (1977) studied the 1972 high school graduates who withdrew from college, but had not reenrolled
within two years. Compared to the 6.2 percent overall unemployment rate of those looking for work from the
class of 1972, the students who withdrew from two-year colleges and those who withdrew for academic
reasons from four-year colleges had relatively high unemployment rates of 9.6 percent. However, those who
withdrew voluntarily for nonacademic reasons had a rate which was only slightly higher than average, 6.5
percent. The four-year college dropouts who had a job were more likely to be disSatisfied with it (21 percent
dissatisfied) than were two-year college degree recipients (14 percent) or two-year college dropouts (16 percent), particularly as an opportunity for using past training and education (41 percent), for promotion and ad·
vancement (35 percent), and for developing new skills (31 percent).
Astin provided data shown in Table 6, by sex and marital status, on employment in 1972 of 1968 college
entrants who dropped out and who graduated.
Employment Status of Dropouts and Graduates Four Years After College
Entrance, by Sex and Marital Status•
Part-time job or
looking for job
Not looking for job
Graduates (not in graduate school)
Part-time job or
looking for job
Not looking for job
• Astin (1975, page 19)
Among the dropouts, married men were the most likely to have a full-time job and married women were
the least likely to have one. Married female dropouts were the most likely not to be looking for a job.
Although married graduates held full-time jobs at about the same rate as married dropouts, the married
graduate sex differences were markedly attenuated. Single graduates were less likely to hold full-time jobs
than single dropouts; in particular, many single male graduates appear to choose to take time off after college.
Comparing the dropouts and graduates for three classes at Princeton University, Parvin et al. (1966, pp.
46-47) concluded that, at least economically, it is not clear that graduates have the advantage. In two of the
three classes, dropouts had slightly higher average incomes than graduates. Although they cautioned readers
not to draw conclusions because of possible bias in the self-selected nature of the returns of questionnaires,
they did conclude that the data indicate many Princeton University dropouts find themselves unable or unwilling to use their talents in the academic world but able to prosper in the business world.
Many researchers conclude that dropping out is overrated as a problem, and that the dropout is usually
taking an assertive step of independence from the parents, a constructive response to a situation that he or
she finds untenable (e.g., Timmons, 1978; Haagen, 1977; Kesselman, 1976; Cope and Hannah, 1975). Most
students who take a leave are happy about it, do return, and graduate at almost the same rate as those who do
not take a leave. A great number who do not return have full-time jobs and do well economically.
Conversely, other researchers, led by one of the foremost and current in this area, Alexander Astin (1975,
1977), suggest that students should be counseled against stopping out. Astin found that students with any interruption in their four years of college resemble dropouts more than persisters. He expressed fears that taking a leave of absence at best delays, and at worst precludes, the usual developmental changes that occur in
college, occupational entry, and subsequent progress.
Astin notwithstanding, we tentatively conclude that students who take a leave of absence and return experience beneficial effects. We also conclude that although students who take a leave of absence and do not
return lose the beneficial effects of college completion, many fare well indeed. There is no evidence that college
completion without the motivation to do so has the beneficial effects that are usually associated with college
STUDENT DEMOGRAPHIC FACTORS
Age. About half of the research studies have found that attrition rates are similar for old and young students
(e.g., Kohen et al., 1978; Gable, 1957; Bragg, 1966), half have found that older students are more likely to
drop out (e.g., Brummer et al., 1978; Astin, 1975; Trent and Medsker, 1968; Newman, 1966).
Sex. Studies conclude that there are no dropout differences between the sexes, that males drop out more, and
that females drop out more. At least five factors explain these conflicting results.
(1) Men are more likely to stop out during their undergraduate years, but are more likely to return and
eventually graduate. Therefore, studies of dropouts during the freshman year are more likely to show that
males have higher dropout rates, and long-term follow-up studies are more likely to show that females have
higher dropout rates. In the College Board's Admissions Testing Program Summary Reporting Service (ATPSRS), colleges that identify which of their enrolling freshmen persist through the freshman year receive profiles that can be used to examine the characteristics of freshman-year dropouts. In the approximately 300 colleges receiving this information, males consistently have a higher freshman-year dropout rate by about two
percentage points, even after controlling for academic abllity of the students and size and academic level of the
colleges. Also, in the 10-year follow-up study reported in El-Khawas and Bisconti (1974), females show a much
higher four-year graduation rate, by about 10 percentage points. However, 10 years after college entry, males
have a higher graduation rate by about five percentage points!
(2) Women in four-year colleges are more likely to transfer to another college (Timmons, 1978; Cope and
Hannah, 1976; Bayer et al., 1973). Therefore, studies of dropouts from individual colleges that count transfers
as dropouts are likely to show higher female dropout rates than are institutional studies that exclude
transfers or studies of dropouts from higher education in general. Also, men are more likely to transfer from a
two-year to a four-year college (Bayer et al., 1973).
(3) In the last 10 years, the number of women entering higher education has increased rapidly. Since
women entrants used to be a more selective group, older studies that did not control for academic abllity
would have shown relatively lower female dropout rates than newer studies or studies that did control for
(4) Because women tend to leave for nonscholastic reasons and men are more likely to be academic
dropouts (Pantages and Creedon, 1978; Tinto, 1975; Spady, 1970), studies that focus only on voluntary
withdrawals show women with relatively higher dropout rates than studies that do not distinguish between
voluntary and involuntary withdrawals.
(6) There are large differences between the sexes at different institutions: for example, women are more
likely to drop out when the ratio of men to women is high (Astin, 1977; Cope et al., 1971; Astin, 1964) and
men are more likely to drop out of large, nonselective universities (ATP-SRS data; Astin, 1975).
Parental Income. ATP-SRS freshman year dropout rates are about the same for all income levels except that
the very lowest levels, below $9,000, exhibit somewhat higher dropout rates. Astin (1976) shows more pronounced four-year dropout rate decreases as income increases, among all income levels. However, the inclusion of income in a regression analysis with parental education, student ability, and motivation failed to add
anything over and above the contribution of the other variables.
Parental Education. Although Pantages and Creedon (1978) conclude that parental education does not appear
to be a major factor, the preponderance of evidence is that it is (e.g., Fetters, 1977; Astin, 1976). The student
from an educated family is more likely to value higher education, even after controlling for the effects of other
Father's Occupation. Some students give lower dropout rates for professional occupational levels (e.g., Trent
and Medsker, 1968), and others do not (e.g., Rossmann and Kirk, 1970). However, the dropout-rate differences among occupational levels disappear when high school performance is controlled (Summerskill,
Ethnic Group. Without controls for academic ability and socioeconomic status, blacks have slightly higher
dropout rates than whites. However, after controls, most researchers have found that the rates are essentially
the same (e.g., Kohen et al., 1978; Fetters, 1977; Astin, 1975); in fact, Peng and Fetters (1978) found that the
dropout rate for whites was slightly higher than for blacks. Astin (1976) found that blacks are more likely to
drop out from predominantly white colleges than from predominantly black colleges. Fetters (1977) found
that after controls Hispanics have lower dropout rates than blacks or whites in four-year colleges; Astin
(1976) found a. lower four-year college dropout rate for Chicanos. Astin (1976) also found that after all other
variables are considered, being Oriental reduces the chance of dropping out by 10 percentage points for
women and 7 percentage points for men.
Religion. With or without controls, Jewish students have been consistently found to be more likely to persist
than non-Jewish students (e.g., Fetters, 1977; Astin, 1976; Newman, 1966).
Marital Status. Astin (1976) reported that although only 2 percent of students were married when they
entered college, marital status had a substantial effect on persistence for both sexes, but in the opposite direction. It decreased the man's chances of dropping out by about 8 percentage points, but it increased the
woman's chances of dropping out by about 11 percentage points.
Hometown Location. ATP-SRS data consistently show that whereas the highest dropout rates are for those
out-of-state students who are from states that are not contiguous to the state in which the college is located,
the lowest dropout rates are for students from contiguous states; typically, dropout rates of in-state students
are in the middle.
Hometown Size. Students from small towns appear to be somewhat more likely to drop out (Astin, 1976;
Cope, 1972; Newman, 1966). There is no consistent pattern for big cities or surburbs.
STUDENT ACADEMIC FACTORS
High School Record. High school grade-point average (GPA) and high school class rank are the best predictors of persistence and attrition (e.g., Pantages and Creedon, 1978; Fetters, 1977; Astin, 1976; Summerskill,
1962). The correlations of high school record with persistence have ranged from 0.26 to 0.60. In a recent
multicollege four-year study, Astin (1977, page 108) shows a correlation of 0.29, out of a multiple correlation
of 0.42 for all freshman characteristics used.
Scholastic Aptitude Test (SAT). ATP-SRS data show that the freshman year dropout rate ranges from 9 percent for those scoring 600 or above on SAT-mathematical sections to 27 percent for those scoring below 300.
The difference in four-year dropout rates probably would be greater. However, after controlling for high
school record, the additional predictive strength is only moderate.
College Board Achievement Tests. ATP-SRS data show that for relatively selective colleges (SAT-verbal score
average of at least 480) the freshman-year dropout rate of students who merely took an Achievement Test is
half that of those who did not ta.ke a.n Achievement Test. Also, the higher the scores, the less likely is the student to drop out.
High School Program. A college preparatory program is a. major factor in predicting retention (Peng a.nd Fetters, 1978; Fetters, 1977). Of course, this ma.y be both because this type of program attracts students who are
more likely to persist and because it prepares students to be better qualified for and have a greater commitment to college education.
High School Academic Rating. Astin (1976) has found that student ratings of the academic quality of their
high schools add significantly to the precision of the dropout prediction. This contrasts with earlier unsuccessful attempts to use the academic quality of the high school to adjust high school grades in order to predict
college grades (Lindquist, 1963).
High School Size. Some studies have found that graduates of very small high schools are more likely to drop
out (Cope, 1972; DeVecchio, 1972; Little, 1969), but others have found no such relationship (Panos a.nd Astin,
1968; Schmid and Reed, 1966; Slocum, 1966).
High School Type. Research findings are mixed (Pantages and Creedon, 1978; Sexton, 1966). ATP-SRS data
show that the freshman-year dropout rates between public a.nd private high school students a.re almost identical. If there is any difference, it is in nonselective colleges, where private high school students are slightly
more likely to drop out.
Years of Study of Various Subjects. ATP-SRS data show that students who study English (particularly),
mathematics, foreign language, and physical science for more years in high school a.re more likely to persist
through their freshman year of college, even after controlling for selectivity of college. However, the number
of years of study of social studies or biological science is unrelated to persistence.
Performance in College. Most studies have found a very significant relationship between performance in
college and attrition, even after other variables are controlled. The causation may be in either direction,
though. Poor grades may be the result of a. decision to drop out, or poor grades may be the cause of a "failure
identity." Astin (1976) goes so far as to suggest more college grade inflation in order to reduce dropout rates.
Areas of Study. After controlling for college academic level, ATP-SRS data show that some areas of study have
relatively high dropout rates-agriculture, forestry and conservation, physical education, architecture and environmental design, engineering, psychology, and police science-and others have relatively low dropout
rates-biological science, health-related professions (particularly premedicine), history and cultures, and
elementary education. However, the most interesting aspect of these data is the differences among colleges,
perhaps indicating the need for college-specific dropout information.
Study Habits. Generally, students with poor study habits are more likely to drop out, and those who spend
more time per week studying are more likely to stay (Pantages and Creedon, 1978; Demitroff, 1974; Sexton,
1966). Astin (1976), however, finds two results which, on the surface, are surprising. (1) Students admitting
to "careless mistakes on tests" were more likely to persist: perhaps "careless mistakes" are more likely to be
made by the better student or the student who has the abillty to be self-critical. (2) Students who "kept my
desk or study place neat" or who "carefully went over diagrams or tables in the textbook" tended to drop out:
perhaps "neatness" means doing little or no homework and "carefully" is a.n attempt to go over minute
details to comprehend them.
STUDENT MOTIVATIONAL FACTORS
Student motivational factors may be considered the sine qua non of persistence, and therefore the most important target of persistence research. However, they may be considered so obviously related to persistence as to
make research on the relationship trivial.
With the exception of those who do not have the reqUisite ability, students continue in college because
they choose to do so a.nd drop out because they choose to do so, for reasons that may or may not be accurately
assessed. For a student with the requisite abillty, even involuntary withdrawal due to low grades is really voluntary: a result of the student's choice not to do the work that is necessary to obtain sufficiently high grades.
Therefore, motivational factors are really the same as persistence factors: the student who wants to do so
will persist and the student who wants to do so will drop out. Any college, parental, or peer group interaction
has the potential to affect the student's satisfaction or dissatisfaction with college, and therefore his or her
motivation to continue.
Of course, the student's precollege attitude may be a very strong influence. For some students, expectations may be self-fulfilling prophecies: some students who are not highly committed toward a degree, who expect to dislike the college, and who expect to leave before obtaining a degree, do so; some students who are
highly committed toward a degree, who expect to like the college, and who expect to obtain a degree, do so.
However, unfortunately confounding the large number of research studies that have analyzed motivational
factors is a well-known opposite effect: that high expectations (perhaps unrealistically high) can lead to disappointment, and that low expectations (perhaps unrealistically low) can lead to a pleasantly surprising satisfaction. Also, although commitment is usually helpful, a compulsion to achieve may lead to, or be a symptom of,
psychological stress, which in turn could lead to dissatisfaction with the college, poor performance, and dropping out.
The vast amount of research on motivational factors is not much help in analyzing how these factors affect persistence, as can be seen in the following sections.
Degree-Level Goal. ATP-SRS data show that, even after the controlling for college ab111ty level, students who
do not expect to obtain a four-year degree are much more likely to drop out during their freshman year. Fetters (1977) confirmed this result; he showed two-year nonacademic dropout rates from four-year colleges of
54 percent for students who did not expect to enter college, 41 percent for students who expected to enter college but not to complete it, and only 19 percent for students who expected to complete college. However, except for those who do not expect to complete a four-year degree, ATP-SRS data show no differences in first-year
dropout rates among students who do or do not want graduate study and who are or are not decided on their
educational goal. However, with respect to four-year dropout rates, Astin (1976) finds that, after controlling
for other factors, compared with students who aspire toward a doctoral or professional degree, wanting a
master's degree adds about 6 percentage points to the dropout rate and wanting a terminal bachelor's degree
adds about 12 percentage points to the dropout rate.
Vocational Goal. Some researchers have found that having a vocational goal increases persistence (Frank
and Kirk, 1976; Hanson and Taylor, 1970; Sexton, 1966), but others have found no such relationship (Panos
and Astin, 1968; Schmid and Reed, 1966; Slocum, 1966). ATP-SRS data show that being undecided on area of
study is not related to a higher first-year dropout rate.
Precollege Dropout Expectation. Several findings have supported the "self-fulfilling prophecy" theory (Pantages and Creedon, 1978; Rossmann and Kirk, 1970; Marks, 1967). However, although Astin (1976) found
that the four-year dropout rate for students who said they had a very good chance of dropping out was twice
that of students who said they had no chance of dropping out, after other variables were controlled, the remaining effect of precollege expectation was very small.
Reasons for Attending College. Iffert (1968) found no relationship between dropouts and persisters with
respect to the reasons students originally gave for attending college.
Parental Influence. Most research has shown that parental expectations of college completion are strong
determinants of student goals and student persistence (see, for example, Tinto, 1975; Sexton, 1966). However,
some studies have failed to find this relationship (Rossmann and Kirk, 1970; Barger and Hall, 1966). Pantages
and Creedon (1978) conjecture that parental influence depends on the quality of the parent-student relationship. Timmons (1978) concludes that dropping out is often a constructive step of the assertion of independence
from parents, especially for males who withdraw voluntarily.
STUDENT PERSONAL CHARACTERISTIC FACTORS
Personality Traits. It is possible to find studies that conclude dropouts are autonomous, mature, intellectual,
committed, creative, assertive, critical, nonconforming, and unconventional; it is possible to find other studies
that conclude dropouts are irresponsible, anxious, impulsive, rebellious, unstable, unimaginative, unadaptable, aloof, disagreeable, immature, impetuous, self-centered, uncooperative, likely to overemphasize per-
sonal pleasures, resentful of regulations, lacking self-sufficiency, and uncertain about the future (Pantages
and Creedon, 1978; Cope and Hannah, 1975). After Sharp and Chason (1978) found several Minnesota
Multiphasic Inventory (MMPI) scales to be variables that significantly distinguished between dropouts and
persisters, they tested their results on a second group of students. The results failed to replicate, and they concluded that prior research showing significant relationships between personality traits and persistence were
Expressed Need for Personal Counseling. ATP-SRS data show consistent findings for 1976 and 1977 that
students who wish to receive personal counseling assistance are 8 percentage points more likely to drop out of
a relatively nonselective college in their freshman year than students who do not seek personal counseling
assistance. It is very likely that this difference would be even greater with respect to a longer term dropout
rate. In contrast, the freshman year personal counseling dropout rate difference was consistently only 2
percentage points for relatively selective colleges.
Smoking Cigarettes. A surprising finding of Astin's (1975) study is that, after controlling for other variables,
smoking has a pronounced relationship with attrition, especially for black students: a black student who
smokes has a dropout probability which is about 20 percent greater than one who does not smoke (the comparable difference for whites is 6 percent).
GENERAL COLLEGE ENVIRONMENTAL FACTORS
Type of College.
Two-year colleges have much higher two-year withdrawal rates than four-year colleges: 39
percent compared to 24 percent (Fetters, 1977, page 23). These differences are primarily, but not completely,
explained by greater student dropout-proneness, including lower ability level, lower student degree expectations, and fewer institutional resources for financial aid, job opportunities, and student housing (Astin, 1975).
Control of College. ATP-SRS data show that even if academic level, student ability, and parental income are
controlled, public four-year colleges have a freshman dropout rate that is about 3 percentage points higher
than that of private four-year colleges (including universities). Fetters (1977) shows a 6 percentage point
public-private difference after the second year. Astin (1975) shows a four-year dropout rate difference of 3
percentage points, primarily because of the difference for males in universities (males in public universities
had the highest dropout rate of 33 percent and males in private universities had the lowest dropout rate of 20
percent). Astin (1975) points out that private colleges have lower dropout rates than public colleges because of
greater resources in residence halls, financiall!.id, and work opportunities. Astin (1977) argues that the expansion of the public sector of education, with its reduced facilities and opportunities for student involvement, in effect dilutes the benefits of college, in addition to increasing the dropout rate. ATP-SRS data show
that for small colleges, where the differences in resources may not be significant, public colleges have lower
Religion. Religiously affiliated colleges, especially those associated with Roman Catholicism, have substantial
holding power over what would be predicted from student and institutional characteristics (Astin, 1976).
Perhaps this is because of an emphasis on community and cohesiveness.
Coeducation. Dropout rates at single-sex colleges are substantially lower. For men, this difference is
accounted for by student and institutional variables. For women, other variables only partially explain this
difference: perhaps it is because women at single-sex institutions have less of a tendency to marry and are
more involved and satisfied with college (Astin, 1976, 1977).
Location. With or without controlling for student or institutional factors, colleges in the Northeast and South
have relatively low dropout rates, and colleges in the West and Southwest have relatively high dropout rates
(Astin, 1975, 1977).
Academic Level. Fetters (1977) shows that the higher the academic level, based on admissions-test score
averages, the lower is the dropout rate. ATP-SRS data show that this is true even if individual test scores are
controlled; that is, for any given student test-score level, the dropout rate is higher at institutions that have
lower test-score averages. Attending a selective college results in a lower undergraduate grade-point average,
but also results in much higher student satisfaction with college and does not increase the chances of drop
ping out or going on to graduate school (Astin, 1977). Most of these effects are due to the increased opportunities for financial aid, work status, and residence at selective colleges.
Size. For both 1977 and 1978, ATP-SRS data show higher dropout rates for large colleges that are nonselective or moderately selective, particularly for males with low test scores. Some researchers have also found
higher dropout rates for large colleges (Holmstrom and Knepper, 1976; Nelson, 1966), others have found the
opposite (Tibby et al., 1978; Kamens, 1971), and still others have found that size has no significant or clear effect (Pantages and Creedon, 1978; Astin, 1975; Tinto, 1975). In his large survey of students entering in 1966,
Astin (1977) found that almost all student involvement, which he finds is the most important aspect of persistence, is decreased by attending a large rather than a small institution-the only exception was involvement in student demonstrations.
Match of Student and College. Different colleges have different images and therefore attract different types
of students. The "college fit" theory states that the greater the congruence between the student's background,
values, goals, attitudes, and interests, and those of most of the students at the college, the more likely it is that
the student will persist. In almost all research studies, this theory has been supported (Pantages and
Creedon, 1978; Starr et al., 1972; Hackman and Dysinger, 1970). Astin (1975) specifically evaluates certain
types of fits. In particular, he finds that the fit theory has a substantial effect on persistence for size of
hometown and size of college, and moderately high effects for religion, race, and the comparison of parental
education with selectivity level of the college. However, he rejects the notion that there are beneficial per·
sistence effects if the student's ability is matched with the selectivity level of the college. ATP-SRS data confirm
this result: students with high test scores are more likely to persist than students with low test scores, for any
college selectivity level; and colleges with high selectivity levels have lower dropout rates than colleges with
low selectivity levels, for any student test-score group. To the extent that one accepts the fit theory, one could
try to identify and track, for special counseling or for increasing their involvement with campus activities,
students who are different from typical students at the college.
COLLEGE PROGRAMS AND THEIR EFFECTS
A college cannot, and, for the benefit of its students, ought not try to hold on to every student who wishes to
drop out or stop out. Indeed, the ideal retention program may not be a retention program at all. Although it
may be possible to reverse some inclinations to leave, a more complete and long-term beneficial strategy is to
encourage retention of students inclined to leave, of current students not yet inclined to leave, and of future
students, by upgrading the level of educational service that is provided for the benefit of the students.
In this age of concern about enrollments, a college should resist the temptation to recruit as many students as
possible, in any way that works. Only by undertaking a higher social responsibility to help students enter the
postsecondary course of study that would best meet their needs can the college help to reduce the personal, institutional, and social costs of an academic mismatch.
One way of promoting student satisfaction with the college is to ensure that from the beginning-that is,
before enrollment-the student receives information about the college that is as accurate, up to date, complete,
and understandable as possible. Although this applies to every institutional-student contact, the primary
source of information about the college is the college catalog. Seagle (1969) found major discrepancies between
catalog descriptions and perceptions of enrolled upperclassmen at eight colleges. Some of the discrepancies
were out-of-date information, course descriptions that are significantly different from course content, pictures
of students working on scientific equipment not ordinarily available to them, and many pictures of minority
students when very few minority students were enrolled at the colleges.
Incomplete catalog information is even a bigger problem. A survey of 210 catalogs by the Carnegie Council
on Policy Studies in Higher Education (1979) discovered that 81 percent did not list the instructors of various
courses, 72 percent did not tell when or how often courses were offered, 18 percent had no information on
high school courses required for admission, 17 percent had no information on financial aid, and 5 percent did
not describe the content of the courses. Unless specific information is provided otherwise, it is very possible
for the potential student to expect that all courses are available every year, that all courses are taught by
faculty members, that all faculty members teach undergraduates, and that no particular high school courses
are required. A student with some of these misperceptions may become disillusioned with the college and later
drop out. What may be a small, insignificant disappointment if known before enrollment can become a major
source of dissatisfaction if discovered after enrollment.
There may be another cause for concern about catalog accuracy and completeness. Mancuso (1978)
reports that some courts are now recognizing the student right to sue for breach of contract on the ground
that services were not as promised in the catalog and were therefore defective.
Colleges should consider providing all information deemed relevant for an informed decision to purchase
the educational service the college provides, including accurate cost projections, expected student performance, current student and faculty perceptions, a complete description of student services, a description of
student life, a description of all academic options on and off campus, student attrition rates, regional and national information on the availability of jobs by career field, and the assessment by graduates of the relationship between their education and job requirements. Stark and Marchese (1977) describe an action taken by
Barat College, and suggest that it be made a common practice: an external audit of the college catalog for both
accuracy and completeness.
In addition to its role in providing potential students with accurate and complete information, the admissions
office is also in an ideal position to initiate a student data base and to coordinate the flow of information
throughout the institution that is necessary for academic and support services. Usually before an application
is made, both the College Board's Admissions Testing Program (ATP) and the American College Testing (ACT)
program supply, in addition to admissions test scores, such a great variety of information about the student
that the ATP or ACT report could be the foundation of a student data base. The ATP report, for example, provides Scholastic Aptitude Test, Test of Standard Written English (TSWE), and Achievement Test scores and information about high school courses, performance, and graduation date; ethnic group; citizenship; whether
English is the best language; veteran status; degree-level goal; college attendance plans (full-time or part-time,
day or evening); plans to ask for educational, vocational, career, or personal counseling; plans to ask for
special assistance in mathematics, reading, writing, or study skills; intended field of study (first and second
choice among 222 specific fields); plans to seek advanced placement, by subject area; self-reported skills and
abilities in 14 different areas; housing preferences; extent of part-time work in high school and needs for college; extracurricular activities in 8 different areas in high school and plans for college; and special information about transfer students. Information in this report could be used to inform students of specific college
programs, to counsel students, and to place students in courses. This information could be augmented first by
additional information supplied on the admissions application and then as events take place during the student's undergraduate years.
Compiling this information in a systematic way is a major step. But it is not sufficient. The information
must flow at the right time, in individual or summary form, to those responsible for assessment, course placement, basic skills, tutoring, counseling, financial aid, housing, and student activities. One important type of
use would be to communicate the existence of services or activities to those students who most likely will
benefit from them. Some examples are specific types of special assistance for those who have indicated a need
or have given themselves low self-ratings, opportunities for credit-by-examination, or specific extracurricular
activities. The admissions office should coordinate this data flow.
In terms of recruitment, other than the obvious advice to find students who are as capable and motivated
as possible, the general evidence that has been compiled on attrition is not very helpful. However, two small
nuggets of information may be useful for planning recruitment strategies:
(1) Students from contiguous states are surprisingly persistent.
(2) There is a positive effect on persistence if the relative size of the college corresponds to the relative
size of the student's home town: small colleges are more likely to retain students from small towns
and large universities are more likely to retain students from big cities.
College Costs and Ways to Meet Them
Financial difficulty is the second most frequently cited category of reasons for dropping out, but after controlling for academic ability and motivation, there is almost no relationship between income and attrition.
Because of the availability of financial aid, there is a general consensus among researchers that financial problems alone do not cause a student to drop out (Kesselman, 1976). However, various financial factors could
have important motivational effects.
Tuition. Astin's (1975) excellent research of 41,000 students at 358 colleges provides a good deal of information of the effects of various financial factors. After controlling for student ability, motivation, income, financial aid received, work status, and residence, tuition did not have any effect on persistence, with one exception: blacks tended to drop out of predominantly black colleges at a significantly higher rate if the tuition was
Parental Aid. In his earlier study, Astin (1975, page 52) reported that parental aid had a moderate effect on
persistence, reducing the four-year dropout rate by about 3 percentage points. He found this effect to be very
large for blacks at predominantly white colleges (17 percentage points), but very small for blacks at
predominantly black colleges (1 percentage point). In his later study, Astin (1977, page 109) reported an
overall parental aid effect that was twice as large (6 percentage points).
Spouse Support. For students who were married before entering college, Astin (1975) found that major
spouse support (more than 50 percent) had a dramatic effect on persistence chances (28 percentage points for
men and 15 percentage points for women). However, minor spouse support (less than 50 percent) had a
negative effect that was worse than no support at all.
Financial Aid. Reported results of research on the effect of financial aid on persistence is particularly confusing. In analyzing National Longitudinal Study data for the class of 1972 by means of a log-linear model,
Fetters (1977, pages 87, 88) concludes that "financial aid was a significant variable in relation to withdrawal
from the four-year college," particularly for students with low income and high aspirations. However, in
analyzing the same data, at about the same time, by means of a multiple regression model, Pang and Fetters
(1978, page 367) conclude that "neither scholarship nor loans have a significant relationship to college
withdrawal in either the four-year or two-year institutions."
Scholarships or Grants. Astin (1973) at first concluded that scholarships or grants increase chances of persistence substantially (on the order of 10 to 15 percentage points), but later concluded (Astin, 1975) that they
have very little effect (about 2 percentage points). Some other researchers agree with the initial finding (e.g.,
Selby, 1973; Blanchfield, 1971), but other researchers agree with the latter finding (e.g., Kohen et al., 1978;
Fields and LeMay, 1973; Iffert, 1958).
Loans. Blanchfield (1971) found that loans did not provide incentive for the student to persist. Astin (1975),
in fact, found that for men the effect is actually negative: a freshman male increases his freshman year
dropout rate by 6 percentage points by receiving a loan.
Employment. Researchers have consistently found that full-time employment has a significantly negative
effect on freshman persistence (e.g., Kohen et al., 1978; Astin, 1975; Iffert, 1958). However, Kohen et al.
(1978) found that among juniors and seniors this effect was no longer present.
Kohen et al. (1978) also found that part-time employment has a significant negative effect on freshman
persistence, although the extent of the effect is much less than for full-time employment. However, Astin
(1975) is quite emphatic about the extensive positive effects of campus involvement resulting from part-time
employment on campus (if the student receives no grant or loan support). He goes so far as to make a rather
unusual, but interesting, suggestion that students eventually take over some of the nonacademic functions of
the institution through part-time work, including maintenance, buildings and grounds, clerical work, food
service, student bookstore, etc.
Other Forms of Financial Aid.
Astin (1975) concludes that
-Federal work-study programs enhance persistence, particularly among blacks and students from
-Personal savings are not significantly related to persistence,
-GI benefits have a negative impact on persistence, perhaps because veterans may find it difficult to adjust to college life,
-ROTC benefits have a very positive effect on persistence, probably because of the commitment represented by participation in ROTC,
-Most financial aid packages appear to have negative effects on persistence: either a work-study opportunity or a grant would be superior to both.
Information. Another important financial factor that is related to persistence is the extent and quality of
financial aid information supplied to students. In a survey of New Jersey college-going students (New Jersey
Commission on Financing Postsecondary Education, 1976), more than one-quarter of the students with parentalincomes below $6,000 failed to apply for financial aid, and 40 percent of those in the $6,000-12,000 range
also failed to apply. When asked, two-thirds of these students said they did not have financial aid information.
A report by the College Scholarship Service (CSS) of the College Board (1977) suggests that this lack of information is not necessarily the fault of the students. It finds that college publications are very unclear and
incomplete, partly because they are poorly written and partly becauE'Ie information is withheld on purpose for
any of the following reasons: full costs will scare students away, the information is too complicated, there are
already too many financial aid applicants, and policies are purposely kept unclear to avoid public scrutiny. As
a result, many students may be deprived of the motivational enhancement of financial aid.
CSS also reported that a survey of students shows they do want long-range information on costs (fouryear, year-by-year, how and when they are to be paid, and the chances and probable extent of future increases) and detailed information on financial aid (kinds available, jobs available, procedures for applying,
and likelihood of obtaining the different types of aid). The CSS report suggests a common format for providing
cost and financial aid information. A good follow-up study would be to evaluate the relationship between good
financial aid information, student success in obtaining financial aid, and persistence.
Obviously, to retain students by helping them make the most of their college experience, a good start is
necessary. Therefore, the orientation program takes on an especially important role.
For this purpose, several suggestions have been made to upgrade the orientation process. Noel (1978) suggests that the college supply "survival tools" in question-and-answer format dealing with common problems
and services that are available to students. He also suggests that the college could make a special effort to
bring together students with common interests. Figler (1979) suggests that orientation is an important time to
supply career information, including data on past graduates.
Astin (1975) suggests that the orientation session should be an information exchange on persistence. The
college could inform the students of college-specific data on prior dropouts (including different student
characteristics and activities) and of some of the generalized findings reported here. A worksheet whereby the
student could calculate his or her chances of dropping out, and see how these chances would change as a
result of some of the choices to be made, could be quite informative. At the same time, the students could supply information to the college to be used in persistence research, future counseling, and perhaps' in planning
for the next freshman class.
With respect particularly to equal-opportunity students, Garcia and Seligsohn (1978) suggest that the
university make a contract with each student to supply specific support services on a continuing basis (out of
operating funds). In return, the student would contract to keep up his or her academic performance and to
use the services whenever necessary.
Chickering and Hannah (1969) suggest that, because of the importance of parental expectations and support, orientation sessions should be extended to parents. These sessions could inform parents about the
typical pressures and concerns that lead to dropping out and indicate the existence of resources that might
help parents and students make more informed decisions if the need arises.
Several studies have found that the interaction between faculty and students is one of the most important
determinants of retention of students (Pascarella and Terenzini, 1979b; Pantages and Creedon, 1978; Terenzini, 1978). It leads to both academic and social integration in the college (Tinto, 1975). It is related to higher
grades (Pascarella and Terenzini, 1978), greater self-perceived intellectual growth (Pascarella and Terenzini,
1978; Wilson et al., 1976), and higher interpersonal self-esteem (Astin, 1977). Astin (1977) reports that it has
the strongest relationship of any factor to overall student satisfaction with the college experience. Wilson et al.
(1975) confirm this result and report that students who interact more with faculty are sought out more by
their peers, read more books for pleasure, are more likely to have graduate degree plans, and make more progress in specific academic skills, such as the ability to evaluate materials and deal with abstractions and
After controlling for seven preenrollment variables and six postenrollment variables, Pascarella and
Terenzini (1979b) found a correlation of 0.24 between faculty contacts and persistence. Contacts involving in·
tellectual concerns were significant for male and female students. Those involving career concerns and information about courses or academic programs were also significant for males. Contacts having to do with campus issues and informal socializing were significant for females. In a more recent study of the interactive ef·
facts between preenrollment variables and faculty contacts, Pascarella and Terenzini (1979a) found that the
faculty contact effect on persistence operates primarily to compensate for low levels of parental education,
commitment to the goal of college graduation, satisfaction with the quality and impact of peer relationships,
Wilson et al. (1975) conducted a comprehensive analysis of faculty-student interaction, including data
from 1,000 faculty members at six diverse colleges. The authors found that not only is it important to the
students, but both students and faculty identified it as the most important characteristic that distinguishes ef·
fective teaching. However, data presented in this analysis (page 34) show that faculty members interact with
students at the rate of only one per day. In a two-week period, they interact approximately four times in the
role of instructor, four times in the role of educational adviser, two times in the role of friend, two times in the
role of career adviser, one time in the role of counselor, and one time in the role of campus citizen. Of course,
many of these contacts could be repeated ones with the same few students. Also, these are averages: some
faculty members have much more interaction than others, and small colleges seem to foster much more in·
teraction than large institutions, which are more research oriented.
However, as stated by Wilson et al. (1976, page 124):
Institutions of higher learning, like any of society's institutions, will never have enough personable,
dynamic, and magnetic staff or bright, highly motivated, self-confident clients, no matter how diligently
they attempt to recruit them. Such persons are in relatively short supply. Effective college teaching and
learning, however, depend not only on the personal qualities of faculty and students but also on the
nature of the relationships by which they are joined. Since interpersonal relationships are heavily in·
fluenced by institutional arrangements, substantial improvement in education might be achieved by
creating conditions that maximize the likelihood of significant encounters occurring between greater
numbers of teachers and students.
The institution could begin by sponsoring faculty sessions to promote the importance of faculty-student
interaction in effective teaching and learning and in increasing student retention. Data from Wilson et al.
(1976) show that students felt that the faculty members who helped them most did so by being available and
open to any discussion, by intellectually stimulating them, by developing feelings of confidence in the
student's abilities, and by demanding high-quality work. Since both students and faculty agree that continui·
ty of interaction is important, colleges ought to involve faculty members in early freshman activities: orientation, advising, and residence hall programming. They also ought to encourage more detailed feedback to each
student about his or her performance than a letter grade, more availability to students, and more time spent
on campus and in the student's environment.
Although almost all faculty members feel that effectiveness as a teacher should be very important, not
many view it as important in promotion and salary decisions. Instead, they say that research and scholarly
activity, school service, and seniority are usually more important. If faculty-student interaction is to improve,
incentive and reward structures at the college, for both teaching and advising, must encourage it.
One possibility is to utilize student feedback on the quality of instruction, and to relate results to promotion and salary decisions. An excellent tool for this purpose and for the purpose of faculty self-evaluation is
the Student Instructional Report, sponsored by the Educational Testing Service. It not only includes an
analysis of faculty-student interaction, but also of teaching techniques and the organization and content of
Since a college exists primarily to supply academic services, it is expected that the quality of the academic programs offered would be related to student retention. Indeed, Steele (1978) found that student perception of
progress toward an academic and career goal was the most significant correlate of student retention at the
University of Miami, Florida, and student perception of the quality of the faculty was the second most significant correlate.
Several suggestions for the improvement of academic programs follow:
-Notify students early, preferably by means of the catalog, about any unusual features of a program (for
example, if a biology department specializes in developmental biology rather than the general field).
-Plan and advertise program changes well in advance, continuing programs until current students
-Establish needed remedial programs in the summer before college entrance.
-Implement a comprehensive student remedial plan-perhaps including tutoring, programmed instruction, study-skill improvement, and self-paced learning-and make it widely known to students.
-Astin (1975) found that credit by examination is significantly related to student retention, especially
for blacks. It may also be important for enticing stopouts to return.
-The selection of students for honors programs may have substantial motivating effects.
-Student ratings of faculty may be helpful in identifying and correcting instructional weakness. The
Student Instructional Report of the Educational Testing Service has scales for course organization and
planning, faculty-student interaction, communication, course difficulty and workload, textbooks and
readings, and examinations.
-Opportunities for undergraduate research should be expanded, since research involvement strongly affects achievement, grades, persistence, aspirations for advanced degrees, and student satisfaction (Astin,
1977, page 222).
-Astin (1975) tentatively concludes that formal foreign study is related to persistence, but foreign travel
without formal study increases chances of dropping out.
-Since withdrawal typically occurs during a vacation or semester break, after the end-of-semester periods
of stress and anxiety (Pantages and Creedon, 1978; Sharp and Kirk, 1974; Iffert, 1958), reduce the
number of semester breaks.
-Establish and consult departmental student advisory groups for staffing, budgeting, and course scheduling.
Counseling and Advising
More recommendations for reducing attrition involve counseling and advising than any other college program
(Pantages and Creedon, 1978). Most studies have shown substantial beneficial effects. For example, Glennan
(1975) reported that intrusive college counseling reduced the freshman attrition rate at the University of
Nevada at Las Vegas from 45 percent to 6 percent in two years. In a study at the University of California at
Berkeley, Frank and Kirk (1975) found that although students who voluntarily used the counseling center did
not differ from students who did not use the counseling center (in terms of test scores, major fields, vocational
interests, or other personal background characteristics), the five-year dropout rate for counseled students was
only 27 percent, compared to 36 percent for all students, and the more frequent the use the lower was the
Regardless of their ultimate effect on attrition, counseling and advising services are important to help
students attain as much as possible out of their educational experience. Brown et al. (1971) matched experimental and control groups in terms of age, sex, text scores, and high school record at Southwest Texas
State University, San Marcos, and found that the experimental group, which was given counseling, had
significantly better attitudes, study habits, and study skills.
The typical undergraduate is subject to several very strong forces. The student may be away from home
for the first time, subject to a challenging, diverse, and foreign environment. There is a need to fit, to be accepted. At the same time that there is very strong peer pressure, often to resist the "establishment," there are
needs to conform, to obtain adult approval, and to find adult models. There is a need to develop a sense of
responsibility, to learn how to balance short-term and long-term benefits and costs. The need for occupational
development is often fraught with anxiety because society has high expectations, but there usually is little opportunity to sample the options. There are fears of measuring up and of loss of freedom. There is the need to
come to terms with sexual development, to learn how to give and receive love. For many, there are also financial concerns. Perhaps the most important force is the need to develop an identity by taking the difficult step
of asserting independence from parental direction. At this time of conflicting forces and maximum energy,
colleges herd students into large classrooms and ask them to sit still in order to acquire skills that will be
useful later in life. When there is a strong conflict between the student and the environment, students who
have avoidance tendencies may drop out, students who view themselves as adaptive may try to change
themselves, students who view the environment as adapative may try to change it, and students who have a
high tolerance for frustration may tolerate discomforts and persist (Timmons, 1976).
In the midst of these psychodynamics, faculty members are called upon to counsel students. The main
areas of student concern are the following.
Financial: Aid, jobs, and money management.
Academic: Program of study, courses, basic skill development, tutoring, study habits, time management,
and test taking.
Personal adjustment: Housing, student activities, social problems, family problems, handicapped needs,
value and goal clarification, personal awareness and growth, decision making, assertiveness, and dealing with
Career development: Identification and evaluation of options, curriculum planning, experiential learning,
The good counselor must have a. thorough knowledge of the resources and services of the college, must be
able to assess (and even anticipate) student needs accurately, and must be able to match student needs and college services. The counselor must be easily accessible to the student, and the advisory relationship is best
when there is reasonable continuity. Since depression is often associated with the act of dropping out (Beck
and Young, 1978), there is a need to deal with it by means of identifying and working on irrational thoughts
and fears; developing better decision-making procedures, study habits, and time schedules; and intervening
with faculty when necessary to arrange for makeup exams or to extend deadlines.
Crocket (1978) identifies 12 essential organizational elements for a successful advising program:
(1) strong commitment of the institution,
(2) a well-articulated institutional policy of who does what, where, when and how,
(3) institutional recognition and reward for good advising and its consideration in promotion and tenure
(4) careful selection of advisers who show interest in advising,
(5) in-service training of advisers,
(6) preparation of an advising handbook, in looseleaf form, containing up-to-date information an adviser
(7) an advising folder for each student, containing all pertinent information,
(8) institutional outreach to students,
(9) advisers not having more than a reasonable student load,
(10) a good referral system to campus resources and services,
(11) accessibility of advisers,
(12) evaluation of the advising system by students.
Because the cost of a large staff of trained professionals administering counseling services is prohibitive,
it is best for colleges to train faculty and perhaps students for counseling. Faculty advising may be an important initial step in establishing meaningful faculty-student interaction that results in both retention and other
benefits. Graduate students or upperclass undergraduate students may be trained as advisers, perhaps as part
of a work-study program, to act as a "first line of defense" without involving the image of official intervention. In either case, advisers must have adequate training, including counseling skills and techniques, role
playing, course offerings, core requirements, referral sources, job opportunities, and data sources.
As stated previously in the discussion on admissions, the establishment of a student data base and the
flow of information throughout the institution are necessary for academic and support services. The admissions office is in an ideal position to consolidate information received from national testing agencies, the admissions process, and events on campus, and to coordinate data flow throughout the institution, in particular,
to student advisers. The advisory folder for each student should contain all pertinent information, and should
be updated after each counseling session.
There is widespread agreement that advising programs will only work if they are well-publicized, to
students, faculty, and parents, and there is institutional outreach to students. Initially, students having many
of the characteristics identified in the first part of this paper as related to dropping out can be identified as
dropout prone for more extensive counseling services. In addition, any cue that something might be wrong
should trigger a counseling interaction: for example, a report of unsat1sfactory or incomplete work, a reduction in course load (Tibby et al. (1978) found that thiB often occurs before dropping out), poor attendance, requests for transcripts, failure to reapply for financial aid, failure to preregister, change of major, failure to
choose a major after two years, taking an off-campus job, taking a full-time job, or moving off campus. Finally, all students should have a regularly scheduled counseling session at least once a semester. In the case of an
intention to withdraw, an exit interview should be required.
In the National Longitudinal Study of the class of 1972, for academic withdrawals (defined in terms of C
average or less or withdrawal because "courses were too hard" or "fa111ng or not doing as well as I wanted"),
the reason that was by far given most often for withdrawing was that "I wasn't really sure what I wanted to
do.'' This reason was given by 73 percent of the academic withdrawals from four-year colleges and by 81 percent of the academic withdrawals from two-year colleges. In contrast, the percentages for nonacademic
withdrawals were only 46 percent from four-year colleges a.nd 32 percent from two-year colleges. A very
plausible conclusion is that students who do not have even tentative career plans are not motivated to do required course work.
Watkins (1979) describes the career development process established at Doane College as a model that
other liberal arts colleges might emulate. There are six developmental stages.
(1) Preenrollment includes career exploration conferences for high school juniors a.nd seniors, career
advisory manuals and workshops for high school counselors, a.nd the establishment by each prospective student of a record of all relevant life experiences (which 18 later updated after each undergraduate counseling
(2) Career planning includes training advisers on the importance of career planning for skills assessment
and counseling techniques, a. career library with complete, up-to-date career information, and involvement of
(3) Curriculum development includes evaluation of all courses in terms of how the course content assists
students in learning more a. bout their career options and faculty internships to work with professionals who
will be hiring graduates in their field.
(4) Experiential learning includes a student internship program a.nd volunteer, part-time, and summer
(6) Placement includes the development of a plan and the matching of student skills with employers.
(6) Postgraduation benefits include tuition-free courses and alumni job placement.
Astin (1975) emphasizes that the greater the involvement with others at the college, the more likely the student will persist. Specifically, he found that membership in fraternities or sororities and participation in varsity athletics increase persistence rates for both men and women. Fetters (1977) found that satisfaction with
campus social life was significantly related to persistence at four-year colleges, but not two-year colleges. Tinto (1975) reported that several studies have found that participation in extracurricular activities is directly
related to college persistence, and conjectures that they provide social and academic rewards that heighten
commitment to the institution.
However, Summerskill (1962) and Iffert (1958) found that social isolation is not a significant factor. Also,
research findings with respect to extracurricular activities are complicated by the fact that the longer the student stays, the more chance there is to participate. Nevertheless, Terenzini and Pascarella (1977) found that at
Syracuse University persisters were not more heavily involved in extracurricular activities than dropouts.
Despite the fact that research findings on the relationship between campus activities and persistence have
been mixed, the greater the student involvement in campus activities, the more pronounced are the college effects of a more positive self-image, both in an interpersonal context and in terms of competence in a great
variety of areas, and the more satisfied is the student with the college experience. Therefore, it makes sense
for the college to encourage students to be involved in campus activities. One way to enhance this is to
facilitate the flow of information to advisers about current student involvement and existing opportunities for
Astin (1977, page 109) states: "By far the most important environmental characteristic associated with college
persistence is living in a dormitory during the freshman year.'' After controlling for entering student
characteristics and other college environmental factors, living in a dormitory reduced the probability of dropping out by about 12 percentage points, compared to living with parents or in an apartment. Also, dormitory
residents were more likely to aspire to a graduate or professional degree, were more likely to achieve in leadership or athletics, and were more satisifed with their college experience, particularly in the areas of student
friendships, faculty-student relations, and social life.
However, data from national samples of 10,000 college students in 226living groups compiled by Moos
(1979) show that these positive dormitory effects are substantially reduced for women in coed dorms. Women
in coed dorms had a higher freshman-year dropout rate of 9.7 percent, compared to 6.3 percent for men in
coed dorms, 6.8 percent for women in single-sex dorms, and 7.1 percent for men in single-sex dorms. They
also transferred to a different living unit at a much higher rate: 7.0 percent, compared to 2.6 percent for
women in single-sex dorms, 3.2 percent for men in coed dorms, and 4. 7 percent for men in single-sex dorms.
In addition, they visited the student health center more often: 6.4 visits in their freshman year, compared to
4.0 for men in single-sex dorms, 4.8 for men in coed dorms, and 6.1 for women in single-sex dorms.
Moos finds differential characteristics for different types of dormitories. He finds that men's dorms stress
independence, competition, and nonconformist qualities, but have less of an emphasis on intellectual and
cultural activities. Women's dorms promote emotional support, have well established procedures, and are
orderly, but do not stress independence, innovation, student influence on rules, or competitiveness.
Like women's dorms, coed dorms also promote emotional support. However, they also promote, to a
greater extent, characteristics which are not usually promoted in women's dorms, such as innovation, student influence on rules, and involvement in spontaneous social activities and (to a lesser extent) intellectual
and cultural activities. There is much less emphasis on traditional social orientation, such as dating and talking about dating, and on order and organization. There is about the same low level of competition a.s in
Thus, since coed dorms emphasize both involvement in a community-type atmosphere and more individual innovation and deemphasize tradition and order, they provide a more complicated, heterogeneous environment. Because there are also traditional pressures for academic achievement and social conformity, mostly
from parents, the atmosphere may prove stressful, particularly to women.
Astin (1975) has some additional data on other living arrangements. In addition to the 12 percentage point
reduction in dropping out for dormitory students, the even greater involvement in campus life of fraternities
and sororities reduces the dropout rate by an additional 6 percentage points. Although an apartment does not
promote retention as much as a dormitory or a fraternity or sorority, an on-campus apartment is better than
an off-campus apartment. Surprisingly, compared with living with one's parents, an apartment has a substantial positive effect on retention for men, but has a substantial negative effect on retention for women.
The worst two-year pattern of housing, increasing the probability of dropping out by 18 percentage
points, is to live in the dormitory in the first year and then to return home to live with parents in the
sophomore year. A return home may almost be considered an interim stsp before dropping out. It may also be
a sign of financial problems.
In light of the benefits of involvement in campus life, perhaps the current trend toward reduced dormitory construction ought to be reexamined. A challenge for future cost/benefit research on dormitories is to
include data on the numbers of students expected to complete their degrees and to implement their career
Every effort should be made to fill existing dormitory space, perhaps by reducing costs, improving
facilities, or establishing a residence requirement (Astin, 1975, tentatively concluded that the positive dormitory effects exist even if there is a dormitory requirement). It would be a good idea to experiment with, and
assess the effects of, different dormitory environments, including different staffing, programming, student
groupings, roommate assignments, and exposure to upperclassmen. In addition, the various probabilities of
persistence for different student housing decisions should be made known to students and their advisers.
The Withdrawal Procedure
Research has shown that the dropout decision is not usually impulsive: it is the product of much thought over
a considerable period of time. Initial discussions are with friends of the same sex, then parents, and then with
friends of the opposite sex. Communication with faculty or college personnel occurs much later, after the decision is crystallized (Pantages and Creedon, 1978; Cope and Hannah, 1975; Chickering and Hannah, 1969).
Cope and Hannah found that for more than three-quarters of the withdrawals, the final decision was made
during the summer vacation or during a time when college was not in session, after the end-of-semester
periods of stress and anxiety. Iffert (1968) found that the quarter system does indeed lead to higher attrition
rates because it increases the number of potential stopping places.
Eckland (1964) found that the longer the initial period of enrollment, the greater are the chances for
reenrollment and for eventual graduation. Also, the more involvement by college personnel and the more the
student believes that the personnel there are concerned, the more likely is reenrollment at that college.
However, Cope and Hannah report that in a study of more than 1,000 dropouts from 13 colleges, for
almost half the dropouts, the college knew nothing about where the students were going or what they were doing, and for the other half the information was very often incorrect. Even 1f only to increase the probability of
reenrollment, exit interviews should be required.
Even after controlling for the high dropout proneness of entering students in terms of background, ability,
and aspirations, and the lack of financial aid, job opportunities, and student housing, two-year colleges still
have higher dropout rates than expected (Astin, 1975). Their students also are less likely to be involved with
the college, experience a relatively small increase in interpersonal self-esteem, and are less likely to implement
their career plans (Astin, 1977).
Perhaps two-year colleges could simulate the four-year college environment by weekend or week-long
educational retreats, cultural events on campus, improved parking, organized study groups, on-campus
employment opportunities, and improved campus recreational facilities and activities. They could also do
everything possible to smooth the transition to four-year colleges, including joint meetings with upperdivision colleges to work out details and organize tours of potential transfer colleges.
Administration of the Retention Program
As stated originally, the best retention program may not really be a specific retention program: it may be an
effort to upgrade the level of educational service, in its broadest sense, that is provided for the benefit of the
students. However, particularly if the emphasis is on change from old attitudes of taking students for granted
to new attitudes of serving students, effecting this change could be called a "retention program."
Beal and Noel (1979) summarized survey returns from 879 colleges about their retention programs. Of
those having such a program, one of the initial moving forces was most often the President (48 percent of the
time). Because a. retention program is most likely to cut across all areas of the institution, the President's initial (and continuing) involvement seems imperative. Others who were frequently initial moving forces were the
Vice President for Student Affairs (34 percent), the Vice President for Academic Affairs (33 percent), the Admissions Office (30 percent), and Counseling Services (23 percent). (Note that there could be more than one initial force.)
The majority of the colleges (58 percent) had not assigned anyone in particular to coordinate the retention
program. Of those that had, 20 percent of the time the coordinator turned out to be the Dean of Students, but
others frequently given this responsibility were an academic administrator (14 percent), the Director of Admissions (12 percent), a student affairs staff person (10 percent), or the Director of Counseling or a counselor
(8 percent). This person most frequently reported directly to the President (43 percent of the time), which
again makes the most sense, but sometimes instead reported to the Academic Vice President (24 percent) or
the Student Affairs Vice President (15 percent).
Reports on the retention programs were categorized into 15 major action programs. The most frequently
designated one was learning and academic support (24 percent). Other popular major action programs were
advising (13 percent), orientation (13 percent), and early warning systems (12 percent). When the colleges
were asked to indicate how well these programs worked in terms of retention, the highest ranked programs
were new policies and structures, learning and academic support, and early warning systems. It is very interesting that in terms of the overall success of the programs, none of these three were highly ranked: the top
three were peer programs, career assistance, and multiple action programs. When the colleges were asked to
identify any of 20 specific action programs, the most frequently identified action programs that were included
in the retention program were improvement in academic adVising (52 percent), specific orientation activities
(50 percent), and exit interviews (40 percent).
The best first step in forming a retention program may be what Noel (1978) calls a campus health check.
Although a great deal of evidence has been presented here and elsewhere associating many student and college
characteristics with retention or attrition, there are substantial differences among institutions. Therefore, it is
important to do a local dropout study.
One way to accomplish this is to participate in the College Board's Admissions Testing Program Summary
Reporting Service (ATP-SRS). What is referred to as Round 3 is an analysis of the characteristics of a college's
freshman-year persisters and dropouts in terms of test scores, high school rank, socioeconomic
characteristics, and college plans. An additional service which is part of ATP-SRS, called the Admissions Yield
and Retention Report, calculates retention yields for 600 different student groups. A separate graphical plotting service can also be used to represent both admissions and retention yields by student group.
In addition, a more specific, institutionally tailored study should be conducted each year of both dropouts
and persisters. Dropouts could be contacted before they leave, as part of a. good counseling system, in the process of leaving, as part of the exit interview, or after they leave, by mailed questionnaires (first class, with two
or three follow-ups, is best). In addition to finding out why each student is leaving (left) and what the student
will be (is now) doing, it is important to ascertain the student's experiences since entering college, including
academic and career progress, residence, work status, and financial aid. By obtaining this latter information
from a sample of persisters, the experiences of dropouts and persisters can be compared.
The Student Outcomes Information Service (SOIS) is a new service cosponsored by the National Center for
Higher Education Management Systems and the College Board to facilitate local student attrition studies. A
college may obtain from SOIS one of two 19-item questionnaires (one for four-year and one for two-year col-
leges). They contain questions on demographic background, college goals, college attendance, students'
reasons for leaving, evaluation of various college services, and plans for additional education, in addition to
space for optional personal identification, local items, and writ~n comments. SOIS also provides optional processing and analytical services.
Information on student experiences since entering college can also be obtained as part of a comprehensive
program for institutional self-study, evaluation, and planning, by the Institutional Research Program for
Higher Education (IRPHE) of the Educational Testing Service. One of the several instruments offered for this
purpose is called Student Reactions to College (SRC). There is a different 150-item questionnaire for four-year
and two-year colleges. The results are grouped into 19 areas of interest in 4 major dimensions:
(1) Processes of instruction: Quality, form, student-centered, academic performance, grading, studying,
instructor accessibility, and involvement with faculty.
(2) Program planning: Counseling and advising, planning, and programming.
(3) Administrative affairs: Registration and scheduling, library and bookstore, rules and regulations,
and administrative procedures.
(4) Out-of-class activities: Campus climate, organized student activities, help with living problems, and
financial and related problems.
In addition to results for all students, up to five institutionally defined subgroups may be used. These
subgroups may include students who are defined to be dropout prone or who have already dropped out.
Other IRPHE instruments may be related to students' satisfaction with their educational experience and
thereby to student retention. The Institutional Functioning Inventory (IFI) can be used to compare views of
faculty, administrators, and students on how the institution is functioning: 2 of 11 scales are concern for
undergraduate learning and democratic governance. The Institutional Goals Inventory (IGI) can be used to
ascertain what faculty, administrator, and student views are of the institutional goals, as they exist on campus and as they would like them to be: 6 of the 20 goals are academic development, individual personal
development, vocational preparation, democratic governance, intellectual/aesthetic awareness, and offcampus learning.
One final word on the interpretation of student data: since many student characteristics and college experiences are interrelated, it is best to control for important variables. A relatively simple way is to subgroup
the student data, for example in terms of test scores, and to compare results for different subgroups. A better
way is to use stepwise multiple regression analysis, which automatically controls for variables that had
previously entered the regression equation: Astin's analyses (1975 and 1977) were based on this technique.
Other related analyses that may be used are discriminant analysis, path analysis, and the log-linear model.
The latter is a relatively new technique for categorizing all variables and looking at all possible combinations
in a hierarchical process: it was used along with multiple regression by Fetters (1977), and the two results
Overall Graduation and Dropout Rates.
Based on a representative cross section of four-year colleges, 35-40 percent of the entering freshmen graduate
in four years from their college of original entry. An additional30-50 percent graduate after four years, or
from a different college, or both. The remaining 10-35 percent are dropouts who never receive a degree.
Students' Reasons for Dropping Out.
The reasons students give include academic matters, financial difficulties, motivational problems, personal
considerations, dissatisfaction with college, military service, full-time jobs, the expressed need for new, practical, nonacademic experiences, and the lack of initial plans to obtain a degree.
Does College Make a Difference?
Effects of College. The evidence indicates that college does make a. difference in improvement in self-image,
social maturity, interests, competence, and employment.
Leave of Absence. The return rates of students who take a. leave of absence depend on the type of institution:
they range from 20 to 30 percent for junior or community colleges and from 90 to 95 percent for prestigious
private universities. Graduation rates of those who return are almost as high as those who do not take a. leave,
and returning students overwhelmingly report good experiences during their leave. Because there is no
evidence that nonreturning students are worse off than students who persist even though they have the inclination to take a. leave, the general beneficial effects of college do not seem to be enough of a. reason to
discourage students from taking a. leave of absence.
Table 7 summarizes what the last 50 years of retention research have shown about the types of students who
are more likely to persist and the types who are less likely to persist.
Relationship Between Student Characteristics and Persistence
Related to persistence
High school program
High school academic
High school size
High school type
Years of study of
Performance in college
Not related or not clear
High school record
Related to attrition
Out of state but not
Taking test, particularly
Not taking test
More than average in
foreign language, or
Less than average in
foreign language, or
Years of study in social
studies and biological
Areas of study
Reasons for attending
Expressed need for
Scholarships or grants
Other forms of aid
Related to persistance
Related to attrition
Not related or not clear
History and cultures
Social sciences (other
Architecture and environmental design
At predominantly black
At predominantly white
Part-time on campus
Financial aid packages
Part-time off campus
General College Environmental Factors and Student Persistence
In the same way that some students are more likely to persist than others, some general college environmental factors are more conducive to student persistence than are others. This is partly, but not completely,
because certain types of colleges attract students who are more likely to persist. Table 8 shows the general
environmental characteristics of colleges that are more conducive to student persistence and the
characteristics of those that a.re less conducive.
TABLE 8. Relationship Between General College Environmental Factors and Persistence
Related to persistence
Related to attrition
Type of college
Control of college
Match of student
Not related or not clear
Student-college fit in
terms of hometown/
college size, religion,
race, and parental
education/ select! vity
H1gh-ab111ty student at
Lack of student/ college
fit in terms of hometown/ college size,
religion, race, and
Low-ab111ty student at
H1gh-ab111ty student at
Low-ab111ty student at
Student-College Environment Fit. While accepting the evidence that some students are more likely to persist
than others and that some college environments are more conducive to persistence than others, moat research
has concluded that the fit between the student and the college is an important factor. For example, a student
from a small town is more likely to persist at a small college, and a large university is more likely to retain a
student from a large city; a match between the religious affiliation of the college and the student aida retention; a student is more likely to persist at a college if he or she is in the racial majority. However, there is no
evidence that a match between a student's ability and the selectivity level of the college aids persistence.
What Can a College Do to Enhance Retention?
For some dropouts, their source of dissatisfaction with their college would have been present regardless of the
educational service that was provided: they did not really want to go to college, the time was right for a break
in their formal education, their talents could be better used elsewhere, or their original college choice was an
error. For others, a better, more complete educational service offered by their college may have prevented or
reduced the source of their dissatisfaction, may have changed their perception of long-range college benefits,
and may have resulted in retention. For the benefit of this latter group and for the benefit of the perhaps
larger group of dissatisfied students who would not consider withdrawal as an option, a college can deal with
its dropout problem by doing everything it can to upgrade the educational service, in ita broadest sense, that it
provides its students. Such an attack on the root causes that contribute to attrition would benefit all students
and would be an excellent recruitment tactic for future classes.
Preenrollment. Because what may be a small, insignificant disappointment if known before enrollment can
become a major source of dissatisfaction if discovered after enrollment, colleges should consider providing all
information deemed relevant for an informed decision to purchase the educational service they provide, including accurate coat projections, expected student performance, current student and faculty perceptions, a
complete description of student services, a description of student life, a description of all academic options on
and off campus, student attrition rates, regional and national information on the availab111ty of jobs by career
field, and the assessment by graduates of the relationship between their education and job requirements. An
external audit of the college catalog for both accuracy and completeness would be a good way of accomplishing
Admissions. The admissions office is in an ideal position to initiate a student data base and to coordinate the
flow of information throughout the institution that is necessary for academic and support services. In terms
of recruitment strategies, two small nuggets of information may be useful: students from contiguous states
are surprisingly persistent, and small colleges are more likely to retain students from small towns and college
universities are more likely to retain students from large cities.
College Costs and Ways to Meet Them. Fina.ncia.l difficulty is the second most frequently cited ca.tegory of
reasons for dropping out, but after controlling for a.ca.demic a.bllity and motivation, there is almost no relationship between income and attrition. Beca.use of the availabllity of financia.l aid, there is a. general consensus
among researchers that financial problems alone do not ca.use a student to drop out. In fact, financia.l aid may
be a motivational enhancement. Also, the ca.mpus involvement of part-time work on ca.mpus ma.y promote
Students say they want long-range information on costs-four-year, year by year, how and when they are
to be paid, and the chances and probable extent of future increases-and detailed information on financial
aid-the kinds ava.ila.ble, the jobs available, procedures for applying, and the likelihood of various students obtaining the different types of aid. There may very well be relationships among good financial aid information,
student success in obtaining financia.l aid, and persistence.
For a good start of the college experience, the orientation program could
include "survival tools" in question-and-answer format dealing with common problems and services
that are available to students,
bring together students with common interests,
supply career information, including data on past graduates,
inform the students of college-specific data on prior dropouts, for different student characteristics
and activities, and of some of the genera.lized findings reported here,
supply a worksheet whereby students could ca.lculate their chances of dropping out, and see how the
chances would change as a result of some of the choices to be made,
set up a contract with each equal opportunity student to supply specific support services on a continuing basis (out of operating funds) in return for the student's promise to keep up his or her
a.ca.demic performance and to use the services whenever necessary,
be extended to parents, informing them a.bout the typica.l pressures and concerns e.tudents face that
lead to dropping out and to indica.te the existence of resources that might help parents and students
make more informed decisions if the need arises.
Faculty-Student Interaction. Severa.l studies have found that the interaction between faculty and students is
one of the most important determinants of retention of students. However, faculty members interact with
students a.t the rate of only one per day.
Colleges should create conditions that would ma.x1mize the likelihood of significa.nt encounters occurring
between faculty and students, for effective teaching and learning and for increasing student retention by
sponsoring faculty sessions to promote the importance of faculty-student interaction,
involving faculty members in early freshman activities (orientation, advising, and residence ha.ll programming),
encouraging more detailed feedback than a. letter grade to each student about his or her performance,
more a.va.ila.bllity of faculty to students, a.nd more time spent by faculty on ca.mpus and in the student's
establishing incentive and reward structures for fa.culty-student interaction, particularly in faculty
promotion and salary decisions.
Aca.demic Programs. Since a. college exists primarily to supply a.ca.demic services, student perception of programs leading toward a.n a.ca.demic and career goal and student perception of the quality of the faculty are
significant correlates of retention. Some suggestions for improvement are
notify students early of unusual program features or changes,
implement a comprehensive remedial plan, and begin it in the summer before college begins,
encourage credit by examination and the establishment of honors courses,
utllize student ratings to identify and correct instructiona.l weakness,
expand opportunities for undergraduate research,
encourage forma.l foreign study,
establish and consult departmental student advisory groups,
reduce the number of semester breaks.
Counseling and Advising. Most studies show that counseling and advising services have substantial beneficia.l effects on retention, but, regardless of their ultimate effect on retention, they are important to help
students attain as much as possible out of their educational experience. Some important aspects of a successful advising program are
strong institutional commitment,
a well-articulated policy of who does what, where, when, and how, including regularly scheduled
sessions at least once a semester,
institutional recognition and reward, particularly in promotion and tenure decisions,
careful selection of advisers, including both faculty and graduate students, who show interest and
aptitude in advising,
in-service training, including counseling skills and techniques, role playing, course offerings, score
requirements, referral sources, job opportunities, and data. sources,
preparation of an advising handbook, in looseleaf form, containing up-to-date information of the resources and services of the college,
an advising folder for each student, containing all pertinent information from national testing agencies, the admissions process, and events on campus,
advisers who identify and work on irrational thoughts and fears, develop better decision-making procedures, study habits, and time schedules, and intervene With faculty when necessary to arrange for
makeup exams or to extend deadlines,
outreach to students having many of the characteristics that make them more likely to drop out or to
those whose actions indicate that something may be wrong, for example, a report of unsatisfactory or
incomplete work, a reduction in course load, poor attendance, requests for transcripts, failure to reapply for financial aid, failure to preregister, change of major, failure to choose a major after two
years, taking an off-campus job, taking a full-time job, or moving off campus,
a reasonable student load for advisers,
a good referral system to campus resources and services,
accessib111ty of advisers,
evaluation of the advising system by students.
Career Development. Students who do not have even tentative career plans are not motivated to do required
course work. Colleges could
offer preenrollment career exploration conferences,
prepare career advisory manuals,
emphasize to advisers the importance of career planning,
run a complete, up-to-date career library,
evaluate course content in terms of career options,
encourage faculty contact with off-campus professionals,
encourage student experiential learning,
develop a comprehensive student placement plan, before and after graduation.
Campus Activities. Despite the fact that research findings on the relationship between campus activities and
persistence have been mixed, the greater the student involvement in campus activities, the more pronounced
are the college effects of a more positive self-image, both in an interpersonal context and in terms of competence in a great variety of areas, and the more satisfied is the student with the college experience. Therefore,
it makes sense for the college to encourage students to be involved in campus activities. One way to enhance
this is to fac111ta.te the flow of information to advisers about cui'rent student involvement and existing opportunities for involvement.
Housing. Perhaps the most important environmental characteristic associated With college persistence is
living in a dormitory during the freshman year. However, the positive dormitory effects are substantially
reduced for women in coed dorms. The greater involvement in campus life of fraternities and sororities has additional retention effects, above those of freshman-year dormitory residence. Every effort should be made to
fill existing dormitory space-perhaps by reducing costs, improving fac111ties, or establishing a residence requirement. It would be a good idea to experiment with, and assess the effects of, different dormitory en·
v1ronments, including different staffing, programming, student groupings, roommate assignments, and exposure to upperclassmen. In addition, the various probab111ties of persistence for different student housing
decisions should be made known to students and their advisers.
The Withdrawal Procedure.
Even if only to increase the probab111ty of reenrollment, exit interviews should
Two-Year Colleges. Even after controlling for the high dropout-proneness of entering students in terms of
background, ability, and aspirations, and the lack of financial aid, job opportunities, and student housing,
two-year colleges still have higher dropout rates than expected. Their students also are less likely to be
involved with the college, experience a relatively small increase in interpersonal self-esteem, and are less likely to implement their college plans.
Perhaps two-year colleges could simulate the four-year college environment by weekend or week-long
educational retreats, cultural events on campus, improved parking, organized study groups, on-campus
employment opportunities, and improved campus recreational facilities and activities. They could also do
everything possible to smooth the transition to four-year colleges, including joint meetings with upperdivision colleges to work out details and organize tours of potential transfer colleges.
Administration of the Retention Program. The best retention program may not really be a specific retention
program: it may be an effort to upgrade the level of educational service, in its broadest sense, that is provided
for the benefit of the students. However, particularly if the emphasis is on change from old attitudes of taking
students for granted to new attitudes of serving students, effecting this change could be called a retention program. The moving force should be the President, and the coordinator should report directly to the President.
Although a great deal of evidence has been presented here and elsewhere associating many student and
college characteristics with retention and attrition, there are substantial differences among institutions.
Therefore, it is important to begin by doing a local dropout study. The Admissions Testing Program Summary
Reporting Service (ATP-SRS) of the College Board or the Student Outcomes Information Service (SOIS) of the
College Board and the National Center for Higher Education Management Systems may be useful for this purpose. Additional information on student experiences after entering college can be obtained through the many
services of the Institutional Research Program for Higher Education (ffiPHE) of the Educational Testing
Astin, Alexander W., Four Critical Years: Effects of College on Beliefs, Attitudes, and Knowledge. San Francisco: Jossey-Bass, 1977.
Astin, Alexander W., Preventing Students from Dropping Out. San Francisco: Jossey-Bass, 1975.
Astin, Alexander W., "Student Persistence: Some Stay, Some Don't-Why?", College and University, VoL48,
1973, pp. 298-306.
Astin, Alexander W., "Personal and Environmental Factors Associated with College Dropouts Among High
Aptitude Students," Journal of Educational Psychology, VoL 56, 1964, pp. 219-227.
Barger, B., and E. Hall, "Time of Drop Out as a Variable in the Study of College Attrition," College and University, VoL 41, 1965, pp. 84-88.
Bayer, Alan E., Jeannie T. Royer, and Richard M. Webb, Four Years After College. Washington, D.C.: American
Council on Education, 1973.
Beal, Phillp E., and Lee Noel, What Works in Student Retention. American College Testing Program, Iowa City,
Iowa, and National Center for Higher Education Management Systems, Boulder, Colorado, 1979.
Bean, Andrew G., and Robert W. Covert, "Prediction of College Persistence, Withdrawal and Academic Dismiseal: A Discrimination Analysis," Educational and Psychological Measurement, Vol. 33, 1973, pp.
Beck, Aaron T., and Jeffrey E. Young, "College Blues," Psychology Today, VoL 12, September 1978, pp.
Blanchfield, W.C., "College Dropout Identification: A Case Study," Journal of Experimental Education, VoL
40, Winter 1971, pp. 1-4.
Bloom, Bernard L., editor, Psychological Stress in the campus Community. New York: Behavioral Publications,
Borup, Jerry H., "Synthesis of Research on College Dropouts and Guidelines for the Future," Talus, Vol. 1,
1968, pp. 21-36.
Bragg, Emma W., "A Study of Student Withdrawal at W.U.," Journal of Educational Psychology, VoL 47,
1966, pp. 199-202.
Brawer, Florence B., New Perspectives on Personality Development in College Students. San Francisco: JosseyBass, 1973.
Brazziel, W1111am, "Non-Intellectual Predictors of Student Persistence/ Attrition and Performance: Implications for College and University Research and Planning," paper presented at the Annual Forum of the
Association for Institutional Research, Montreal, Que., Canada, 1977.
Brown, William F., Nathan 0. Wehe, Vernon G. Zunker, and Warren L. Haslann, "Effectiveness of Student-toStudent Counseling in the Academic Adjustment of Potential College Dropouts," Journal of Educational
Psychology, VoL 62, 1971.
Brunner, Wayne D., Gene Packwood, and Barry Wilson, "Retention and Attrition: Does it Relate to Student
Goals?' •, paper presented at American Educational Research Association, Toronto, Ont., Canada, 1978,
ERIC No. ED163682.
Carnegie Council on Policy Studies in Higher Education, Fair Practices in Higher Education, Reports and
Responses of Students and Their Colleges in a Period of Intensified Competition for Enrollments. San
Francisco: Jossey-Bass, 1979.
Chickering, A.W., and William Hannah, "The Process of Withdrawal," Liberal Education, Vol. 66, 1969,
College Scholarship Service of the College Board, Making It Count: A Report on a Project to Provide Better
Financial Aid Information to Students. New York: College Board, 1977.
Cope, Robert G., and William Hannah, Revolving College Doors: The Causes and Consequences of Dropping
Out. New York: John Wiley, 1976.
Cope, Robert G., "Are Students More Likely to Drop Out of Large Colleges?", College Student Journal, Vol.
6, 1972, pp. 92-97.
Cope, Robert G., K.G. Pailthrop, D.C. Trapp, M. Skating, and R.G. Hewitt, An Investigation of the Entrance
Characteristics Related to Types of College Dropouts, Washington, D.C.: United States Office of Education,
Crocket, DavidS., "Academic Advising: A Cornerstone of Student Retention," in Lee Noel, editor, Reducing
the Dropout Rate, Vol. 3 of New Directions for Student Services. San Francisco: Jessey-Bass, 1978,
Delworth, Ursula, Marv Moore, Julie Millick, and Patrick Leone, "Training Student Volunteers," Personnel
and Guidance Journal, Vol. 68, 1974, pp. 67-61.
Demitroff, John F., "Student Persistence," College and University, Vol. 49, 1974, pp. 663-667.
DeVecchio, Richard C., "Characteristics of Nonreturning Community College Freshmen," Journal of
College Student Personnel, Vol. 13, 1972, pp. 429-432.
Eckland, Bruce K., "Social Class and College Graduation: Some Misconceptions Corrected," American Journal
of Sociology, VoL 70, 1966, pp. 36-60.
Eckland, Bruce, K., "College Dropouts Who Come Back," Harvard Educational Review, Vol. 30, 1964, pp.
Edington, Robert V., and Fred Gilliard, "RAP: Retention Advisory Program," Journal of College Student Personnel, VoL 19, 1978,pp.472-473.
El-Kha.wa.s, Elaine H., "Solving Problems Through Self-Regulation," Educational Record, VoL 69, 1978,
El-Kha.wa.s, Elaine H., and Ann S. Bisconti, Five and Ten Years After College Entry: 1971 Followup of 1961
and 1966 College Freshmen. Washington, D.C.: American Council on Education, 1974.
Fetters, William B., Withdrawal from Institutions of Higher Education: An Appraisal with Longitudinal Data.
Involving Diverse Institutions, National Longitudinal Study, National Center for Educational Statistics.
Washington, D.C.: United States Office of Education, 1977.
Fields, Charles R., and Morris L. LeMay, "Student Financial Aid: Effects on Educational Decisions and
Academic Achievement," Journal of College Student Personnel, VoL 14, 1973, pp. 426-429.
Figler, Howard E., "Building a Career Foundation," in Ed Watkins, editor, Preparing Liberal Arts Students
for Careers, VoL 6 of New Directions for Education, Work, and Careers. San Francisco: Jessey-Bass, 1979,
Frank, Austin C., and Barbara A. Kirk, "Differences in Outcomes for Users and Nonusers of University
Counseling and Psychiatric Services," Journal of Counseling Psychology, Vol. 22, 1976, pp. 262-268.
Gable, Robert Irwin, "A Study of the Student Dropout Problem at Miami University," Ph.D. dissertation,
Miami University, Oxford, Ohio, 1966, listed in Dissertation Abstracts, Vol. 17, 1967, p. 61.
Garcia., Sandra. A., and Harriet C. Seligsohn, "Undergraduate Black Student Retention Revisited," Educational
Record, Vol. 59, 1978, pp. 156-166.
Glenna.n, Robert E., "Intrusive College Counseling," School Counselor, Vol. 24, 1976, pp. 48-50.
Glenna.n, Robert E., "Intrusive College Counseling," College Student Journal, Vol. 9, 1975, pp. 2-4.
Gray, Robert G., Edmund Jacobson, Sidney S. Micek, Cathleen Patrick, Nancy Renkiewicz, and W1ll1a.m Van
Dusen, Student Outcomes Questionnaires: An Implementation Handbook. Boulder, Colo.: National Center
for Higher Education Management Systems, and New York: The College Board, 1979.
Ha.agen, C. Hess, Venturing Beyond the Campus: Students Who Leave College. Middletown, Conn.: Wesleyan
Hackman, J. Richard, and Wendell S. Dysinger, "Commitment to College a.s a Factor in Student Attrition,"
Sociology of Education, Vol. 43, 1970, pp., 311-324.
Hanson, Gary R., and Ronald G. Taylor, "Interaction of Ability and Personality: Another Look at the Drop-Out
Problem in a.n Institute of Technology," Journal of Counseling Psychology, Vol. 17, 1970, pp. 540-545.
Harvey, James, "Preventing College Dropouts: A Review," Currents '70, No. 3, ERIC No. 043799, 1970.
Holmstrom, Engin Inel, and Paula. R. Knepper, "Four-Year Ba.ccala.ur~ate Completion Rates: A Limited Comparison of Student Success in Private and Public Four-Year Colleges and Universities," American Council
on Education, Washington, D.C., ERIC No. ED127877, 1976.
Iffert, Robert E., and Betty S. Clark, College Applicants, Entrants, Dropouts. Washington, D.C.: United States
Office of Education, 1966.
Iffert, Robert E., Retention and Withdrawal of College Students. Washington, D.C.: United States Department
of Health, Education, and Welfare, 1958.
Jencks, Christopher, Who Gets Ahead?: The Determinants of Economic Success in America.. New York: Basic
Johansson, Charles B., and Jack E. Rossmann, "Persistence a.t a. Liberal Arts College: A Replicated Five-Year
Longitudinal Study," Journal of Counseling Psychology, Vol. 20, 1973, pp. 1-9.
Kamens, David H., "The Effects of College on Student Dropout," Northeastern University, Center for Applied
Social Research, Boston, Mass., 1972, ERIC No. ED068038.
Kamens, David H., "The College Charter and College Size: Effects on Occupational Choice and College
Attrition," Sociology of Education, Vol. 44, 1971.
Kesselman, Judi R., Stopping Out: A Guide to Leaving College and Getting Back In. New York: M. Evans&! Co.,
Kohen, Andrew I., Gilbert Nestel, and Constantine Ka.ma.s, "Factors Affecting Individual Persistence Rates in
Undergraduate Programs," American Education Research Journal, Vol. 15, 1978, pp. 233-252.
Kohen, Andrew I., Gilbert Nestel, and Constantine Kamas, "Success and Failure in College: A New Approach
to Persistence in Undergraduate Programs," Ohio State University, Center for Human Resource Research,
Columbus, Ohio, ERIC No. ED124053, 1976.
Kowalski, Cash, Impact of College on Persisting and Nonpersisting Students. New York: Philosophical
Lindquist, E.F., "An Evaluation of a. Technique for Scaling High School Grades To Improve Prediction of
College Success," Educational and Psychological Measurement, Vol. 23, 1963, pp. 623-646.
Little, J.K., "The Persistence of Academically Talented Youth in University Studies," Educational Record, Vol.
Mancuso, John H., "Legal Rights to Reasonable Rules, Fair Grades, and Quality Courses," in Joan S. Stark,
editor, Promoting Consumer Protection for Students, Vol. 13 of New Directions for Higl1er Education.
San Francisco: Jossey-Bass, 1976, pp. 75-88.
Marks, Edmond, ''Student Perceptions of College Persistence, and Their Intellective, Personality and Performance Correlates," Journal of Educational Psychology, Vol. 58, 1967, pp. 210-221.
Maudal, Gail R., James N. Butcher, and Paul A. Mauger, "Multivariate Study of Personal and Academic
Factors in College Attrition," Journal of Counseling Psychology, Vol. 21, 1974, pp. 560-567.
McNeeley, J.H., College Student Mortality. United States Department of Interior Bulletin, No. 11, 1937.
Middleton, Lorenzo, "With Freshmen Scarcer, Emphasis Shifts To Keeping Present Students," Chronicle of
Higher Education, October 30, 1978, pp. 1,6.
Moore, James W., Edward C. Anderson, and Marvin W. Lynch, "An Equal Opportunity Program Retention
Design," College and University, Vol. 51, 1976, pp. 692-699.
Moos, Rudolph H., Evaluating Educational Environments. San Francisco: Jossey-Bass, 1979.
National Center for Education Statistics, National Longitudinal Study, Tabular Summary of the Third Followup Questionnaire Data. Washington, D.C.: 1978.
National Task Force on Better Information for Student Choice, Better Information for Student Choice: Report
of a National Task Force. Washington, D.C.: American Association for Higher Education, 1978.
Nelson, A. Gordon, "College Characteristics Associated with Freshman Attrition," Personnel and Guidance
Journal, Vol. 44, 1966, pp. 1046-1050.
New Jersey Commission on Financing Postsecondary Education, The Needs and Resources of Undergraduate
Students in Postsecondary Education in the State of New Jersey, 1974-75. Princeton, New Jersey: New
Jersey Commission on Financing Postsecondary Education, 1976.
Newman, M.A., The Student and the College Community: A Study of Attrition and Persistence in a Highly
Selective Liberal Arts College. Chicago: University of Chicago Press, 1965.
Noel, Lee, editor, Reducing the Dropout Rate, Vol. 3 of New Directions for Student Services. San Francisco:
Noel, Lee, "College Student Retention-A Campus-Wide Responsibility," National ACAC Journal, Vol. 21,
Panos, Robert J., and Alexander W. Astin, "Attrition Among College Students," American Educational
Research Journal, Vol. 5, 1968, pp. 57-72.
Pantages, Timothy J., and Carol F. Creedon, "Studies of College Attrition: 1950-1975," Review of Educational
Research, Vol. 48, 1978, pp. 49-101.
Pascarella, Ernest T., and Patrick T. Terenzini, "Interaction Effects in Spady's and Tin to's Conceptual Models
of College Dropout," Sociology of Education, Vol. 52, l979a, pp. 197-210.
Pascarella, Ernest T., and Patrick T. Terenzini, "Student-Faculty Informal Contact and College Persistence:
A Further Investigation," Journal of Educational Research, Vol. 72, l979b, pp. 214-218.
Pascarella, Ernest T., and Patrick T. Terenzini, ''Student-Faculty Informal Relationships and Freshman
Year Educational Outcomes," Journal of Educational Research, Vol. 71, 1978, pp. 183-189.
Pascarella, Ernest T., and Patrick T. Terenzini, "Patterns of Student-Faculty Informal Interaction Beyond
the Classroom and Voluntary Freshman Attrition," Journal of Higher Education, Vol. 48, 1977,
Pascarella, Ernest T., and Patrick T. Terenzini, ''Informal Interaction with Faculty and Freshman Ratings
of Academic and Nonacademic Experience of College," Journal of Educational Research, Vol. 70,
1976, pp. 35-41.
Patrick, Cathleen, Edwin Myers, and William Van Dusen, A Manual for Conducting Student Attrition Studies.
Boulder, Colo.: National Center for Higher Education Management Systems and New York: The College
Board, revised edition 1979.
Pedrini, Bonnie C., and D.T. Pedrini, "Evaluating Experimental and Control Programs for Attrition/Persistence," Journal of Educational Research, Vol. 71, 1978, pp. 234-237.
Peng, SamuelS., and William B. Fetters, "Variables Involved in Withdrawal During the First Two Years of
College: Preliminary Findings of the National Longitudinal Study of the High School Class of 1972,"
American Educational Research Journal, Vol. 15, 1978, pp. 361-372.
Parvin, Lawrence A., Louis E. Reik, and Willard Dalrymple, editors, The College Dropout and the Utilization
of Talent. Princeton, New Jersey: Princeton University Press, 1966.
Ramist, Leonard, ''Admissions-Yield and Persistence Analysis,'' paper presented at the Annual Forum of the
Association for Institutional Research, Houston, Texas, 1978.
Rever, Phillp R., and Lawrence K. Kojaku, "Access, Attrition, Test Scores, and Grades of College Entrants and
Persisters: 1965-1973," Research in Higher Education, Vol. 5, 1976, pp. 289-299.
Rossmann, Jack E., and Barbara A. Kirk, "Factors Related to Persistence and Withdrawal Among University
Students," Journal of Counseling Psychology, Vol. 17, 1970, pp. 56-62.
Rubin, Howard S., and Helen A. Cohen, "Group Counseling and Remediation: A Two-Faceted Intervention
Approach to the Problem of Attrition in Nursing Education," Journal of Educational Research, Vol. 67,
1974, pp. 195-198.
Savicki, Victor, Harry Schumer, and Robert E. Stanfield, "Student Role Orientations and College Dropouts,"
Journal of Counseling Psychology, Vol. 17, 1970, pp. 559-566.
Scher, Murray, "Small College Deals with Student Retention," Journal of College Student Personnel, Vol. 20,
1979, p. 176.
Schmedinghoff, Gerard J., "College Program for High-Risk Students," College and University, Vol. 54, 1979,
Schmid, John, and Stanley R. Reed, "Factors in Retention of Residence Hall Freshmen," Journal of Experimental Education, Vol. 35, Fal11966, pp. 28-35.
Sedlacek, William E., and Dennis W. Webster, "Admission and Retention of Minority Students in Large
Universities," Journal of College Student Personnel, Vol. 19, 1978, pp. 242-248.
Selby, James E., "Relationships Existing among Race, Student Financial Aid, and Persistence in College,"
Journal of College Student Personnel, Vol. 14, 1973, pp. 38-40.
Sexton, Virginia Staudt, "Factors Contributing to Attrition in College Populations: 25 Years of Research,"
Journal of General Psychology, Vol. 72, 1965, pp. 301-326.
Sharp, Lawrence F., and L. Ralph Chason, "Use of Moderator Variables in Predicting College Student
Attrition," Journal of College Student Personnel, Vol. 19, 1978, pp. 388-393.
Sharp, W. Harry, and Barbara. A. Kirk, "A Longitudinal Study of Who Seeks Counseling When," Journal of
Counseling Psychology, Vol. 21, 1974, pp. 43-50.
Sherman, Roger H., and Vincent Tinto, "The Effectiveness of Secondary and Higher Education Intervention
Programs: Critical Review of Research," paper presented at Annual Conference of the American Educational Research Association, Washington, D.C., ERIC No. ED106378, 1975.
Shulman, Carol Hernstadt, ''Recent Trends in Student Retention,'' Research Currents, American Association
for Higher Education, ERIC No. ED127841, Ma.y 1976, pp. 3-6.
Slocum, W.L., "Social Factors Involved in Academic Mortality," College a.nd University, Vol. 32, 1956, pp.
Spady, William G., "Dropout from Higher Education: Toward an Empirical Model," Interchange, Vol. 2, No.3,
1971, pp. 38-62.
Spady, W1ll1a.m G., "Dropouts from Higher Education: An Interdisciplinary Review a.nd Synthesis," Inter·
change, Vol. 1, No. 1, 1970, pp. 64-85.
-Speegle, John, "College Catalogs: An Investigation of the Congruence of Catalog Descriptions of College En·
vironments with Students' Perceptions of the Same Environment as Revealed by the College
Characteristics Index,'' Ph.D. dissertation, Syracuse, University, Syracuse, New York, 1969.
Stark, Joa.n S., The Ma.ny Faces of Educational Consumerism. Lexington, Mass.: Lexington Books, 1977.
Stark, Joa.n S., editor, Promoting Consumer Protection for Students, Vol. 13 of New Directions for Higher
Education. San Francisco: Jossey-Bass, 1976.
Stark, Joa.n S., and Theodore J. Marchese, "Verifying Consumer Information," in Joa.n S. Stark, The Ma.ny
Faces of Educational Consumerism. Lexington, Mass.: Lexington Books, 1977, pp. 179-185.
Starr, Ann, Ellen L. Betz, and John Menne, "Differences in College Student Satisfaction: Academic Dropouts,
Nonacademic Dropouts, a.nd Nondropouts," Journal of Counseling Psychology, Vol. 19, 1972,
Steele, Ma.tt W., "Correla.tss of Undergraduate Retention at the University of Miami," Journal of College
Student Personnel, Vol. 19, 1978, pp. 349-352.
Summerskill, John, "Dropouts from College," in Nevitt Sanford, editor, The American College. New York:
John Wiley, 1962, pp. 627-665.
Terenzini, Patrick T., ''An Evaluation of Three Basic Designs for Studying Attrition,'' paper presented a.t the
Annual Forum of the Association for Institutional Research, Houston, Texas, ERIC No. ED161358, 1978.
Terenzini, Patrick T., a.nd Ernest T. Pascarella, "The Relation of Students' Precollege Characteristics a.nd
Freshman Year Experience to Voluntary Attrition," Research in Higher Education, Vol. 9, 1978, pp.
Terenzini, Patrick T., a.nd Ernest T. Pascarella, "Voluntary Freshman Attrition and Patterns of Social a.nd
Academic Integration in a. University: A Test of a. Conceptual Model," Research in Higher Education, Vol.
Thackrey, Russell I., "If You're Confused About Higher Education Statistics, Remember: So Are the People
Who Produce Them," Phi Delta. Kappa.n, Vol. 56, 1975, pp. 415-419.
Thayer, Robert E., "Do Low Grades Ca.use College Students To Give Up?", Journal of Experimental Education,
Vol. 41, Spring 1973, pp. 71-73.
Tibby, Edward, Judith Bonnett Hira.ba.ya.shi, John Olson, a.nd Richard Peterson, What Really Happened to
the Class of '75?: A College Experience Survey. Berkeley, Calif.: Educational Testing Service, 1978.
'Iimmons, Frank R., ''Freshman Withdrawal from College: Positive Step Toward Identity Formation,''
Journal of Youth a.nd Adolescence, Vol. 7, 1978, pp. 159-173.
Timmons, Frank R., "Freshman Withdrawal from College: An Empirical Examination of the Usefulness of
'Autopsy Studies' , "Psychological Reports, Vol. 41, 1977, pp. 672-674.
Timmons, Frank R., "Research on College Dropouts: A Critical Look," in Bernard L. Bloom, editor,
Psychological Stress in the Campus Community. New York: Behavioral Publications, 1975, pp. 99-123.
Tinto, Vincent, "Dropout from Higher Education: A Theoretical Synthesis of Recent Research," Review of
Educational Research, Vol. 45, 1975, pp. 89-125.
Trent, James W., and Leland L. Medsker, Beyond High School. San Francisco: Jossey-Bass, 1968.
United States Bureau of the Census, Statistical Abstract of the United States. Washington, D.C.: 1978.
Waterman, Alan S., Patricia S. Geary, and Caroline K. Waterman, "Longitudinal Study of Changes in Ego
Identity Status from Freshman Year to Senior Year at College," Developmental Psychology, Vol. 10,
1974, pp. 387-392.
Watkins, Ed, editor, Preparing Liberal Arts Students for Careers, Vol. 6 of New Directions for Education,
Work, and Careers. San Francisco: Jossey-Bass, 1979.
Wessel, Tyrus R., Kenneth Engle, and Uld1s Smidchens, "Reducing Attrition on the College Campus,"
NASPA Journal, Vol. 16, 1978, pp. 26-32.
Wilson, Robert C., Jerry G. Gaff, Evelyn R. Dienst, Lynn Wood, and James L. Bavry, College Professors and
Their Impact on Students. New York: John Wiley, 1975.
Withey, Steven B., editor, A Degree and What Else?: Correlates and Consequences of a College Education.
New York: McGraw-Hill, 1971.